The Spider Genus Latrodectus (Araneae, Theriididae) - Herbert W. Levi, 1959.

The Spider Genus Latrodectus (Araneae, Theridiidae)
Author(s): Herbert W. Levi

Source: Transactions of the American Microscopical Society, Vol. 78, No. 1 (Jan., 1959), pp. 7-
43
Published by: Wiley on behalf of American Microscopical Society
Stable URL: http://www.jstor.org/stable/3223799 .
Accessed: 30/09/2013 17:48
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
Wiley and American Microscopical Society are collaborating with JSTOR to digitize, preserve and extend
access to Transactions of the American Microscopical Society.
http://www.jstor.org
This content downloaded from 129.72.2.27 on Mon, 30 Sep 2013 17:48:37 PM

All use subject to JSTOR Terms and Conditions
MILESTONES IN SOCIETY HISTORY 7
LITERATURECITED
ACKERT, J. E. 1946. Dr. Henry Baldwin Ward (1865-1945). Trans. Amer. Microsc.
Soc., 65: 177-180.
BLACKHAM,G. E. 1882. The evolution of the modern microscope. Proceed.
Amer. Soc. Microscopists, 1882: 25-43.
1884. Memoir of Robert B. Tolles. Ibid., 1884: 41-44.
Cox, J. D. 1884. Annual address of the president. Robert B. Tolles and the
angular aperture question. Ibid., 1884: 5-39.
GALLOWAY, T. W. 1910. The future of this society. Trans. Amer. Microsc. Soc.,
29: 79-84.
1915. The American Microscopical Society. Advertising booklet. Decatur,
Illinois. 10 p.
KRAUSS, W. C. 1901. The debt of American Microscopy to Spencer and Tolles.
Trans. Amer. Microsc. Soc., 23: 19-30. (Portraits).
SMITH,H. L. 1882. Memoir of Charles A. Spencer. Proceed. Amer. Soc. Micro-
scopists, 1882: 49-74.
THE SPIDER GENUS LATRODECTUS

(ARANEAE, THERIDIIDAE)
HERBERTW. LEVI
Museum of Comparative Zoology, Harvard University
The poisonous black-widow spiders1are notoriously difficult taxonomic-

ally and, because arachnologists have been loath to study the group
intensively, there is considerable confusion. It is particularly unfortunate
in a medically interesting group to find physiological observations based on
specimens of uncertain determination (e.g. papers in Buckley and Porges,
1956). It is hoped that some of the facts presented here will stimulate
further collecting of Latrodectusin those areas from which few specimens
are now available, and further study of Latrodectus biology directed
toward the solution of taxonomic problems.
This revision was made possible by a grant from the National Institutes
of Health (No. E-1944). I extend sincere thanks to all those who helped
me: Prof. E. Mayr for advice on taxonomic problems and for reading
the manuscript; Dr. W. J. Gertsch for reading the manuscript; my wife
for editing the paper; Mrs. Nancy Buffler for help with diagrams and
maps; Mrs. Karen Hudner and Mrs. F. Patricia Scher for checking
literature; Miss Helen Vaititis for Russian translation. The work re-
quired large collections made available to me by: Dr. W. J. Gertsch
of the American Museum of Natural History and by Dr. O. Kraus,
Senckenberg Museum. Dr. W. Crome sent the Dahl collection from
the Berlin Museum. Other collections were made available by: Prof.
C. J. D. Brown, Montana State College; Mr. E. Browning, British Museum
(Natural History); Dr. J. Camin, Chicago Academy of Sciences; Prof.
1'Black-widow" is the common name for the genus Latrodectus in the United
States. Each part of the world has its own name for these spiders.

8 HERBERT W. LEVI
H. Dietrich, Cornell University Collection; Mrs. D. Frizzell (Dr. Harriet

Exline); Prof. V. V. Hickman, Hobart, Tasmania; Miss Beverly Holloway;
Dr. A. Holm, Uppsala University Collection; Dr. B. J. Kaston; Dr. R. F.
Lawrence, Natal Museum; Dr. D. C. Lowrie; Dr. B. Main, University
of Western Australia; Mr. J. E. H. Martin, Science Service, Department of
Agriculture, Canada; Dr. G. Noland; Miss B. de Pikelin and Prof. R. D.
Schiapelli, Museo Argentino de Ciencias Naturales; Prof. J. G. Rempel;
Mr. V. Roth; Prof. A. Shulov, Hebrew University, Jerusalem; Dr. P.
W. Smith, Illinois Natural History Survey; Dr. R. H. Smithers, National
Museum of Southern Rhodesia; Prof. M. Vachon, Museum National
d' Histoire Naturelle, Paris; Prof. H. K. Wallace; Mr. G. B. Wiggins,
Royal Ontario Museum; and many others who sent single specimens.
METHODSUSED
Since immatures cannot be identified with certainty, this study has
been limited to sexually mature individuals. Extreme variability in
shape and slope of the clypeus made it necessary to measure carapace
length from the anterior median eyes to the posterior edge of the carapace.
Figures for leg length indicate the length of the first patella-tibia divided
by the length of the carapace. The abdominal pattern drawings are
freehand and stylized. The drawings of genitalia and setae were made
under a dissecting microscope equipped with a calibrated grid in the
ocular. The same enlargement was used for all setae drawings. The
setae illustrated were from the dorsum of the abdomen, slightly anterior
and lateral. In most instances the ectal view of the palpus was drawn
from slightly anterior, looking down at the coil of the embolus. Unless
indicated otherwise, the left palpi are illustrated.
In Latrodectus, as in some other theridiid genera (e.g. Achaeranea,
Theridiotis), females can be determined to species only after examination
of the internal genitalia, as the shape of the epigynum is variable within
species, and is similar among species. The technique for examination
of the internal genitalia is to remove the abdomen from the prosoma,
cut along one side of the epigynum and bend this structure back, then
submerge the whole abdomen in clove oil. An alternate method is to
insert a small hook at the side of the sclerotized area of the epigynum
near the genital groove, and lift the epigynum. Extra tissue is then
cleaned off with a fine needle (a mounted sharpened minuten nadel).
The bibliography lists only the most important papers cited in the
text. References mentioned in distribution records will be found in
Bonnet's Bibliographia Araneorum Vol. 1, and in reference to publications
since 1938, in the Arachnida section of the Zoological Record. Of the
several hundred references consulted, many papers on this genus were
written by biologists unaquainted with spiders and frequently with
little knowledge of taxonomy. Such works have been hard to evaluate
and many of them are not cited. Most American literature had to be
ignored because species had been confused. Although a few are cited,
some West African records are doubtful, since L. geometricus is often
black in this region, whereas L. mactans may be light in color, the reverse
of the situation elsewhere. Some literature records were so general as to
be usless; others were published in unobtainable journals. The distri-
bution records are not complete; but border records are cited. A Russian

SPIDER GENUS LATRODECTUS 9
0 O E
._' Oo
; .:
J O
_J a
MAP 1. Distribution of Latrodectus geometricus.

10 HERBERT W. LEVI
book, P. J. Marikovski, 1956, [Lycosa and Lactrodectus, Morphology,

Biology, and Toxicity] Kirgistan State Akad. Science, Frunze, was seen
only after this manuscript had been completed.
SPECIESCRITERIA
Although the structure of genitalia has been useful in separating
spider species, this criterion in the past has been ignored consistently in
Latrodectus. This may be due at least in part to the striking variation
in coloration and abdominal setae within many populations, tempting
classifiers to use these characters rather than examine microscopically the
slight but critical differences in the genitalia.
The only previous revision of Latrodectushas resulted in controversy.
In 1902 F. P. Cambridge revised Latrodectus. In the same year Dahl
(1902) published on Latrodectus, applying a taxonomic procedure un-
fortunately still in use by some invertebrate zoologists. Dahl very
minutely examined populations from different areas of the world, mounting
cephalothorax, legs, pieces of abdomen and genitalia on slides, and found
that each population differed. Dahl gave species rank to those populations
which differed considerably in leg spination, and gave subspecies rank to
those which differed less. In a short note later in the year, F. P. Cam-
bridge (1902a) commented: "Whether the character deduced from the
differences in the hairs of the patella and tibia will bear the weight of im-
portance laid upon them can only be judged by a comparative examination
of a large series." Dahl was resentful, and in a paper entitled "Can a
taxonomist work too carefully?", (1902a), attacked Cambridge personally.
"In German scientific circles this question would be quite boring, but I
have to raise it since some foreigners find it an ideal to describe many
species very superficially and, it may sound unbelievable, reproach those
who work more carefully, if they make conclusions from their careful
studies." When Cambridge criticized Dahl's disregard of type species
of genera, Dahl was scornful and dismissed Cambridge as an "ignoramus".
In another part of the paper Dahl observed: "Herr Cambridge insists
that hairs and setae of a certain shape which I use systematically vary
between one and six and seven. I do not know how Mr. Cambridge can
make this statement. I did not say so and in nature such a thing does
not occur. .. ." The disagreement exemplifies the "polytypic species
versus typological species" dispute. Cambridge's views are corroborated.
by the present study.
Fifty years ago when only a few specimens were available for study
a simple criterion was employed to delimit species: if it appeared dif-
OF FIGURES
DESCRIPTION
FIG. 1 Latrodectus mactans, left palpus, subventral view, expanded. C, con-
ductor; E, embolus; M, median apophysis; P, paracymbial hook; R,
radix; S, subtegulum; T, tegulum; Y, cymbium.
FIGS. 2-4. L. curacaviensis, palpus (Florida). 2. Mesal view. 3. Ventral view.
4. Subectal view.
FIGS. 5-7. L. mactans, palpus (Virginia). 5. Mesal view. 6. Ventral view. 7.
Subectal view.
FIGS. 8-10. L. geometricus, palpus (Brazil). 8. Mesal view. 9. Ventral view.
10. Subventral view.

ty I mA\
3 kw_24
C>
5 6 7
Uaf
K
9 10

12 HERBERT W. LEVI
ferent it was described as new. At that time it was noted that the
genitalia of spiders (the second pair of appendages, or palpi, of the male;
and the sclerotized genital opening, or epigynum, of the female) differ
in shape in different species. For some time it was thought that only
the degree of difference of the genitalia should be used as species dis-
tinguishing characters. Some authors, studying variable samples, de-
scribed the extremes and discarded intermediates that did not fit their
preconceived ideas of species. Other authors, on finding a specimen
different, but not different "enough" to be a species, described it as a
new "subspecies".
Present day arachnologists, unlike their predecessors, have large
collections available for study. Often large series are available from
different geographic regions. Revisions based on large series invariably
show, in spider species as well as in most other groups of animals, that
due to variability, "degree of difference" can not possibly be used as the
only criterion for delimiting species. My own studies of the theridiid
genera Theridion, Achaeranea and Steatoda (Levi 1955, 1957, 1957b) have
shown that many species are polytypic. Holm (1956) has shown that
Erigone arctica (White) (Linyphiidae) in the Arctic can be divided into a
number of distinct subspecies. The difficulty of placing specimens of
Cicurina (Agelenidae) into the numerous species described by Chamberlin
and Ivie (1940) suggests that they dealt with fewer highly variable species.
In a later paper Chamberlin and Ivie (1941) showed similar small dif-
ferences in species of the related Agelenopsis. In Agelenopsis, however,
not only can the sympatric species be placed, but the species are known
to have different habitats also. This study reveals that Latrodectus is
a genus with highly variable characters. The palpi of all males in a
collection of Latrodectus geometricus C. L. Koch from one locality are
illustrated (Figs. 39-45), and comparison of the figures shows that even
the left and right palpi in the same individual may differ sufficiently to
be considered different species by some arachnologists.2
Although occasionally vigorously defended, the species "by-degree-of-
difference" concept involves still other difficulties. What authority is to
decide precisely what degree of difference is the critical one in dealing
with similar species? Should it be those who delve into the minutiae of
setae3, as Dahl did; those "more careful" workers, who can find characters
hardly anybody else can see? It is not surprising that those who deny
the reality of species often adhere to the typological species concept
(Grant 1957). If we believe that spider species, like species in other
animal groups, originate from isolated geographical races, we can assume
that the extremes of geographical variation within a single species at times
may show greater difference than separate species. Abundant evidence
demonstrates this. Steatoda pulcher (Keyserling); Enoplognatha joshua
Chamberlin and Ivie; Theridion ohlerti Thorell, and Latrodectusmactans
20ne cause for such great differences is that once expanded for copulation, the
parts may not fit back into the palpus as before. Further, weakly sclerotized palpi
may become warped, probably as a result of dehydration in alcohol. Figures 54, 55
show the cymbium of L. mactans (Fabricius) drawn from only very slightly different
angles, to demonstrate that only a minute warping might show as a considerable
difference in an illustration.
3Contrary to the situation in Drosophila, for instance, setation in Latrodectus
appears to have no particular taxonomic significance.

0 E
: ;'
= ..
00e
c
00
CI oo
E " e
=0a
X??
MAP 2. Distribution of Latrodectus mactans.

14 HERBERT W. LEVI
(Fabricius) appear to be species with marked geographic variation. On

the other hand Chrysso albomaculata0. P. Cambridge and Chryssovexabilis
Keyserling (Levi 1957a), although distinct only in coloration and some
obscure genitalic details, must be considered species, since they live in
the same region (sympatric) and seem to be reproductively isolated.
The taxonomist must answer the question when specimens come
from separated localities and show minor morphological differences: do
these specimens represent geographical races or separate species? For
the most part there is no answer to this evolutionary problem; this is a
case for arbitrary judgement. In spiders the number of specimens
available usually is inadequate, and additional series collections eventually
may give the answer. The erroneous claim that because they show
differences, populations "must" be new species, was discussed by Jordan
over 50 years ago when he urged that variable allopatric populations be
considered subspecies. Dahl mistook the differences in his Latrodectus
collections to be indications of species status. Keegan (1945), in listing
the species of Latrodectus,thought he could separate the species by color
pattern.
Modern systematists must incorporate a broad understanding of
genetics, physiology, embryology, comparative morphology, indeed of all
areas of biology, into the taxonomic approach to evolution. Failure to
do so is retrogression to the stage of "pebble-sorting" (Mayr 1957).
While many taxonomists are severely critical of those experimental
biologists who disregard advances in allied fields, the critics themselves
often are guilty of the same narrowness. In dealing taxonomically with
the highly variable species of Latrodectus, we can employ one of three
standards. Like Dahl, we can coin as many new specific names as there
are collections. Or, we can disregard geographic variation, and combine
several grossly differing populations in each widespread species. Or
finally, in line with recent developments in genetics and evolution (and
with the findings of F. P. Cambridge over 50 years ago), we can recognize
polytypic species in Latrodectus.
In a series of specimens from a single locality, the number of species
usually can be determined with some certainty. Concordance of dif-
ferences in several characters of sympatric specimens indicates the presence,
of more than one species, even though the differences may be smaller
than in collections of a single species from several localities. A careful
study of Florida Latrodectus shows that there are three forms. One,
called Latrodectus mactans var. bishopi by Kaston [=L. curacaviensis],
not only has different proportions and coloration, but has a distinctly
different habitat, living well above the ground. Compared with the
other Florida species, it also has consistent differences in genitalia (Kaston
remarked on the similarity of genitalic structure in bishopi and mactans,
presumably because he compared bishopi with specimens from Connecticut
which he thought were mactans, but which actually were also bishopi).
The other two Florida species also differ in details of genitalic structure,
often in coloration, and probably in habits. Using as a yardstick the
structural differences of Florida species, we can examine specimens from
other parts of America for corresponding differences. (It is unfortunate
that mention of habitat characteristics, of so much importance in Latro-
dectus, is often omitted from collection data). Contrary to expectation,

SPIDER GENUS LA TRODECTUS 15
two of the Florida species are found to be sympatric over a large part of
the Americas, but because of strikingly parallel geographic variations in
color, the minor consistent differences in habitat selection and in mor-
phology of genitalia have been overlooked. The species involved are
L. mactans and L. curacaviensis (Miiller) [=bishopi Kaston]. Analysis
of another character, leg length (Fig. 38), reveals different averages for
the two species in different parts of their ranges. All indications suggest
different habits for the two: in Florida, as mentioned above, L. cura-
caviensis lives in trees and shrubs, L. mactans on the ground; at the
northern limits of their range, L. curacaviensis lives under stones and
logs and L. mactans lives on or near buildings and in trash.
Specific morphlogical differences in the length of the embolus of the
male palpus are seen most easily in freshly molted males, with more
difficulty in mated individuals. After mating, the parts of the embolus
do not fit back as before, and length is more difficult to judge. Further-
more, as Dahl pointed out, the embolus commonly breaks in mating, so
that many males have the embolus broken, and about a third of the
females examined have a portion of the embolus remaining in the ducts.4
The ducts of the female correspond in length to the embolus length.
The tip of the embolus lies in the anterior end of the seminal receptacle
and the width of the embolus often stretches the ducts in which it lies.
Latrodectusgeometricushas the longest embolus, L. mactanshas the embolus
one coil shorter, and L. curacaviensis has it two coils shorter. That these
differences in length are normal rather than due to accidental displacement
of the duct, is clearly evident in the palpi of penultimate males in which
the transparent outer skin reveals the ducts in natural position before
displacement. Smithers (1944) thought that the shape of the palpal
cymbium differed in the two South African species. The cymbium shape
differs between L. mactans (Figs. 53-55) and L. geometricus(Figs. 48-51),
but is similar in L. curacaviensis (Fig. 52) and L. mactans; furthermore
there seems to be considerable geographical variation in this character in
L. geometricus (Figs. 48-51). Unfortunately very few males of L. geo-
metricus were available for examination.
Despite the certainty with which series of sympatric specimens usually
can be placed into species, specimens from Palestine are puzzling. Un-
fortunately very little material was available, compared to more than
a thousand American specimens. The differences between females of
L. revivensis and L. mactans are well within the range of variation of
L. mactans. The males seem to differ only in color. Possibly they
represent two populations of the same species, one population having
been introduced from another area. I have not synonymized the third
species, L. pallidus 0. P. Cambridge, although the only differences seem
to be coloration and possibly texture of the abdomen. I have not seen
any males. Figures by Charitonov suggest that the palpal embolus is
similar in length to that of L. mactans. Shulov's (1940) paper indicated
4The frequency with which the embolus breaks after mating is unique; it has not
been observed in any other theridiid species, although in many species, one occasion-
ally finds portions of an embolus in the female ducts. Perhaps the alleged "husband
eating habit" of black-widow spiders is related to this. Does his inability to with-
draw the embolus make the male prospective prey of his mate? Or perhaps the
embolus breaks because the female preys on the male before the palpus has been
withdrawn.

16 HERBERT W. LEVI
that its habits are different from those of L. mactans. It is probably a

good species.
There are indications of marked ecological differences between species,
a fertile field for future studies. Are there localities where two sym-
patric species compete? Does the absence of L. curacaviensis in the
Greater Antilles, a gap in its distribution, indicate a recent invasion of
L. mactans replacing an endemic species? Bianchi (1944) in a short
paper noted that with the disappearance of a population of L. mactans
in Hawaii (possibly due to the introduction of a hymenopterous parasite
Beaus californicus Pierce), geometricus has become abundant in a vicinity
where it was not previously found.
Many of the forms described in the literature have been given "variety"
or subspecific names. Chamberlin and Ivie (1935) in their revision of
Latrodectus north of Mexico divided L. mactans into three subspecies.
Although these subspecific names have had considerable appeal to non-
arachnologists, their legitimacy is doubtful. The three subspecies are
mactans, hesperus and texanus. Female Latrodectus m. mactans (type
locality Massachusetts, where L. mactans has now been found not to
occur) are stated to differ from L. m. hesperus Chamberlin and Ivie (type
locality Utah) in having a light spot on the abdomen above the spinnerets
and in having shorter legs. The latter character is invalidated by Cham-
berlin and Ivie's own data. The third subspecies was characterized by
large size (contradicted by the specimens I have examined) and by vari-
able coloration, most similar to mactans. In their preoccupation with
these subspecific names, Chamberlin and Ivie apparently did not dis-
tinguish between L. mactans and L. curacaviensis, both of which are
found in Utah.
Although not as much information is available on the Latrodectus
subspecies described by other authors, their validity is similarly in doubt.
All widespread species of the genus show wide variation. Representatives
from neighboring colonies may differ greatly in color or structure. Only
certain easily observed characters have been mapped here to suggest the
vast geographical differences. There are all indications that dines of
OF FIGURES
DESCRIPTION
FIGS. 11-12. Latrodectus dahli, female genitalia. 11. Dorsal view. 12. Ventral
view, cleared.
FIGS. 13-14. L. hystrix, female genitalia. 13. Dorsal view. 14. Ventral view.
FIG. 15. Seminal receptacle of L. mactans (Israel), lateral view.
FIGS. 16-18. L. curacaviensis, female genitalia (Br. Columbia). 16. Lateral view.
17. Dorsal view. 18. Ventral view.
FIGS. 19-21. L. mactans, female genitalia (California). 19. Lateral view. 20.
Dorsal view. 21. Ventral view.
FIGS. 22-24. L. pallidus, female genitalia (Israel). 22. Lateral view. 23. Dorsal
view. 24. Ventral view.
FIGS. 25-27. L. geometricus, female genitalia (Brazil). 25. Lateral view. 26.
Dorsal view. 27. Ventral view.
FIGS. 28-31. Abdomens of females. 28. L. geometricus, dorsal view. 29. L.
hystrix, lateral view. 30. L. hystrix, posterior view. 31. L. pallidus,
lateral view.
FIGS.32-37. Epigyna. 32. L. hystrix. 33. L. pallidus (Israel). 34. L. pallidus
(Iran). 35. L. curacaviensis (Paraguay). 36. L. curacaviensis (Florida).
37. L. geometricus (Australia).

11 12 13 14
6"~i
16 19 25
/^Nm22
'5
21
I 24 2
/I \I8 18/j cY1
I iii I
%
I
I .1
K~K .t -1
V,

18 HERBERT W. LEVI
these characters go in different directions, so that the naming of sub-

species entails the choice of an artificially selected group of characters.
Thickness of legs, comparative length of first and fourth legs or leg seg-
ments, the shape of the clypeus, the carapace, and hairs on legs may be
of importance but were not measured despite considerable observed
variation. Virtually nothing is known regarding variability of physio-
logical characters, although Crome (1956) seems to think that the potency
of L. mactans venom varies in different parts of Europe. The naming
of such subspecies would contribute neither to recognition of entities nor
to the study of variation or evolution, but in these instances could lead
to further nomenclatural confusion.
On the other hand, the confusion resulting from synonymizing well-
known names of allopatric populations might be alleviated by using them
as trinomials: L. mactans tredecimguttatus(Rossi) of Europe and North
Africa, L. mactans cinctus Blackwall of South Africa, L. mactans menavodi
(Vinson) of Madagascar, and L. mactans hasselti Thorell of India to
Australia.
Of considerable interest are the parallel adaptations of species. Four
species occurring in Europe and the Near East all have strong abdominal
spines. They are the endemic L. pallidus, L. hystrix and L. dahli and
also the widespread L. mactans. In the Mediterranean region (Map 4)
the latter species has its strongest abdominal spines but has only weak
abdominal setae in America like the sympatric L. curacaviensis. In
America the two sympatric species L. curacaviensis and L. mactans,
although both showing vast coloration differences in different parts of their
range, often have similar coloration in the same region.
As in other theridiid spiders the eyes of Latrodectus are variable in
size. This variation has been noticed before by Gerschman and Schiapelli
(1943) and Smithers (1944), and is very different from other spider
families where eye-size and relationships often make good generic
characters. Simon (1892) used eye-size to delineate theridiid genera, and
was followed by some other authors (Badcock, 1932), resulting in con-
fusion. Only Simon's taxonomic judgment prevented chaos. All pro-
spective students of theridiid species should be warned against basing
species on a single character, such as eye-size. A combination of characters,
emphasizing the ventral view of the palpus, the epigynum, and the internal
female genitalia, provide reliable criteria in this family.
LATRODECTUS WALCKENAER
Latrodectus Walckenaer, 1805, Tableau des Araneides, p. 81. Type
species: Latrodectus13-decimguttatusRossi (= L. mactans) by subsequent
designation of Latreille, 1810, Considerations generales sur l'ordre naturel
des Animaux, p. 424.
Chacoca Badcock, 1932, Journ. Linnean Soc. London, p. 9. Type
species: Chacoca antherata Badcock (=L. curacaviensis) by subsequent
designation of Bonnet, 1956, Bibliographia Araneorum, 2:1030 NEW
SYNONYMY.
Large theridiid spiders. Lateral eyes separated from each other.
Chelicerae with no teeth. Legs of medium length; first usually longer
than fourth. A distinct comb present on fourth tarsus. Abdomen
usually subglobular. Colulus large. Body of male much smaller than
that of females.

- _B P .
. ._
MAP 3. Distribution of dorsal abdominal pattern of female Latrodectus mactans and
carapace length in mm. Small figures next to large ones indicate the number
of specimens whose measurements were averaged.

20 HERBERT W. LEVI
Epigynum (Figs. 32-37) with a depression wider than long, similar in

all species. Seminal receptacles dumb-bell shaped (Fig. 15), connecting
duct a coil of varying length. Palpal cymbium modified (Figs. 1, 48-55).
Median apophysis (M), radix (R) and conductor (C) present, but the last
structure small. Median apophysis fits against paracymbial hook (P'.
Tegulum (T) highly modified, a coiled tube (as in argiopids) rather than
spherical. Length of embolus (E) and shape of basal portion of radix
diagnostic.
The genus differs from Enoplognatha and Steatoda (which also have a
large colulus) in lacking teeth on the chelicerae, in separation of the
lateral eyes, and in the structure of the genitalia. Unlike other theridiid
spiders, the venom of Latrodectusis poisonous to vertebrates.
Six species are recognized in this genus. Two of the species (L.
mactans, L. geometricus)are world wide in distribution.
KEY TO MALES
la. Embolus of palpus with four coils (Fig. 10) .........................geometricus
lb. Embolus with fewer than four coils .......... .... ....................... .. 2
2a. Embolus of palpus with three coils (Fig. 7) ................... mactans, pallidus
2b. Embolus of palpus with two coils (Fig. 4) ......................... curacaviensis
(and also probably hystrix and dahli)
KEY TO FEMALES
la. Posterior lips of epigynum opening extending on each side (Fig. 37); median
parts of connecting ducts loop back near posterior portion of seminal recepta-
cles (Figs. 26, 27); abdomen often gray, sometimes black...........geometricus
lb. Posterior lip rarely extending on each side beyond opening of epigynum (Figs.
32-36); median parts of connecting ducts loop back near anterior portion or
anterior to sem inal receptacles ........................................... 2
2a. Ducts of genitalia with three or more coils ................................. 5
2b. Ducts of genitalia with two coils .................................. ........ 3
3a. South American ................................................ curacaviensis
3b. Not found in South America .............................................. 4
4a. Abdomen with only short, strong setae; in Yemen ............. ......... hystrix
4a. Abdomen with short and long setae; in Iran to Sokotra ...................dahli
5a. D ucts with four or five coils ............. .............................. 7
5b. D ucts w ith three coils ..................................................... 6
6a. From the Americas .............................................. curacaviensis
6b. From South Africa ................................................... mactans
7a. Coloration mainly white; from Turkmen to North Africa .............. pallidus
7b. Widespread, coloration mainly black in Southern Russia and Mediterranean
countries ..................................................... mactans
Latrodectus geometricus is probably native to Africa. Its coloration
is generalized, similar to that in Achaearanea and in some juveniles of
L. mactans, although in parts of its range (particularly Madagascar) a
large percentage of individuals is black. This species is allegedly less
poisonous than others (Baerg, 1954). Females of L. hystrix and L. dahli,
occurring in Arabia and Iran, have the simplest genitalia; the males of
these species are not known. The genitalia are very similar among all
species of Latrodectus, and very distinct from those of other theridiid
genera. There is not, as Simon (1892, p. 561) suggested, a relationship
between Latrodectus and Dipoena. Latrodectus is close to Steatoda and
Enoplognatha, the other theridiid genera with the colulus prominent.
Dipoena has the colulus reduced, and the genitalic structure of Dipoena
is very different from that of Latrodectus.
A number of taxonomic problems remain to be settled: Are the

Paraguayan and Argentine Latrodectuscuracaviensis with short connecting

ducts a different species and are only those specimens L. curacaviensis
which are determined as L. foliatus Mello-Leitao by South American
arachnologists? Present judgment of synonymizing foliatus was made
since the males collected with the females from both Argentina and
Paraguay could not be distinguished from L. curacaviensis. Large col-
lections may provide an answer. Does Latrodectuspallidus differ morpho-
logically from L. mactans; do Shulov's observations on ecological differences
hold over the whole range? Are there other species as yet not recognized
because of past preoccupation with coloration and spination?
It is urged that only those species be recognized which have been
adequately described in terms of diagnostic characters such as length of
embolus, shape of palpal parts and structure of internal female genitalia.
X Florida
0 N. Centraland Eastern U.S.
(S. Carol.-N. Hamp.-Utah)
V Pacific Coast (N. Calif.-Brt. Col.)
A Pacific Coast (Calif.except northerncounties)
.1! vW V VW w v v
o. O
COOOD 00 00 00 9
C0
0
xx x x x x x x x
A V AAAaAA AA AA A
oo o o9o ocaxcCbo
0 X X xx x x x X
E
1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5
ratioof patella-tibiaI length/carapacelength
FIG. 38. Latrodectus mactans and L. curacaviensis leg length in different parts of
northern portion of American range.
LATRODECTUS GEOMETRICUS C. L. KOCH

Figures 8-10, 25-28, 37, 39-50, 80-83; Map 1
Latrodectus geometricus C. L. Koch, 1841, Die Arachniden, 8: 117,
plate 284, figure 684, 9. Type specimens from Colombia. Roewer,
1942, Katalog der Araneae, 1: 425. Bonnet, 1957, Bibliographia
Araneorum, 2: 2368.
Latrodectuszickzack Karsch, 1878, Zeitschr. Ges. Naturwiss., 51: 311,
plate 8, figure 1, 9. Type specimens from Zanzibar.
LatrodectusobscuriorDahl, 1902, Sitzungsber. Gesell. Naturf. Freunde,
Berlin, p. 41. Syntypes from Madagascar in the Zoologisches Museum,
Berlin. NEW SYNONYMY.
Latrodectus concinnus 0. P. Cambridge, 1904, Ann. South African
Museum, 3: 152, plate 10, figures 3, 4, c, 9. Female syntypes from
Capetown, males from Devil's Mountain, South Africa.
Latrodectus geometricus subalbicans Caporiacco, 1949, Comm. Pont.
Acad. Sci., 13: 376. Female from Mackinnon road, 80 km. west of
Mombassa, Kenya. NEW SYNONYMY.

22 HERBERT W. LEVI
Latrodectus geometricus modestus Caporiacco, 1949, Ibid., 13: 376.

Females from Kebete and Nairobi, Kenya. NEW SYNONYMY.
Latrodectus geometricus obscuratus Caporiacco, 1949, Ibid., 13: 377.
Female and male types from Nairobi, Kenya. NEW SYNONYMY.
Comments: Some of Dahl's specimens of L. obscuriorwere examined;
they are dark as are most specimens from Madagascar. Caporiacco's
"subspecies" are color varieties based on dorsal coloration of the abdomen.
Latrodectusg. geometricusCaporiacco corresponds to figures 7a and 7b of
F. P. Cambridge, 1902, subalbicans to figure 7, obscuratus to figures 7d
and 7e. Latrodectus g. modestus has a pattern like that of L. pallidus,
but spination as in L. geometricus.
DIAGNOSIS. The epigynum (Fig. 37) varies little, the lower lip pro-
jects on the sides beyond the upper lip of the opening. The median
portions of the connecting ducts are shorter (Figs. 26, 27) than in other
species. The dorsal abdominal coloration is often spotted gray (Fig. 28),
sometimes black.
The embolus of the male palpus (Fig. 8-10) is longer than that of
L. mactans, having four coils.
VARIATION. Black females are found in all parts of the range, in-
cluding Florida, but the majority of Madagascar specimens examined
are black. There are some differences in the cymbium of the male
palpus (Figs. 48-51). The palpus of South African males is much smaller
in size than those from Australia and South America.
HABITAT. Found on houses and buildings in Florida, and in a similar
habitat in South Africa (Smithers, 1944) where they were found "in a
ventilator, behind pipes, under the eaves, in fact any off corner which
affords substantial shelter for the nest and where a plentiful supply of
insects is to be found."
DISTRIBUTION. Cosmotropical, but most widespread in Africa and
probably native there, although Smithers believed it to be introduced in
South Africa. Many records are probably based on faulty determinations.
California records cited in many papers were first published by Marx
(1890b) whose collections are notorious for their erroneous locality labels.
The record from Utah (Chamberlin and Wood, 1929) was based on an
immature specimen, either L. mactans or L. curacaviensis.
RECORDS. United States. Florida: Charlotte Co.: Punta Gorda (S. Rounds).
Dade Co.: Coral Gables (J. F. Pierson; H. Lewis; Weber); Miami (W. Schoenherr).
Orange Co.: Winter Park (T. C. Lawrence). Palm Beach Co.: Lake Worth.
DESCRIPTION OF FIGURES
FIGS. 39-47. Variation in Latrodectus geometricus palpus. 39-45. From same

locality (Minas Geraes, Brazil). 39-40. Left palpus, specimen A.
41. Mirror image of right palpus, of specimen A. 42-43. Left palpus.
specimen B. 44. Mirror image of right palpus, of specimen B. 45.
Left palpus, specimen C. 46-47. Specimen from Southwest Africa,
46. Left palpus. 47. Mirror image of right palpus of same specimen.
FIGS. 48-51. Geographical variation of cymbium of Latrodectus geometricus palpus.
48. (Minas Geraes, Brazil). 49. (South Africa). 50. (Southwest
Africa). 51. (Darwin, Australia).
FIG. 52. Cymbium of L. curacaviensis.
FIGS. 53-55. Cymbium of L. mactans. 53. (Colorado). 54. (Israel). 55. Drawn
from slightly different angle (Corsica).

-~-- 48

24 HERBERT W. LEVI
Volusia Co.: Daytona Beach (Webber). Mexico. Veracruz: Veracruz (N. L. H.

Krauss). Panama: Panama City (N. L. H. Krauss); Ancon (N. L. H. Krauss); Colon
(Hamilton); Ft. Amador (T. C. Lawrence); Quarry Heights (A. M. Chickering);
Ft. Randolph (A. M. Chickering). Bermuda: (Banks, 1902d). Brit. West Indies:
North Bimini (M. Bishop). Jamaica: Kingston; Story Hill; St. Andrew Par. (A.
Wiles); St. Ann. Par. (R. Bengry'. Navassa Isl. (W. Clench, W. Schevill). Cuba:
Santiago de las Vegas (Baker). Haiti: Damien (H. L. Dozier). Puerto Rico:
Aguirre (Leonard); Rio Piedras (Barker). Trinidad: Navy Base, S. W. Trinidad
(R. Ingle). Aruba: (R. C. Carrell). Curacao: (Hasselt, 1887d). Colombia: (C.
L. Koch, 1841); Barranquilla (N. L. H. Krauss). Venezuela: San Felix, Bolivar;
Caripito (W. Beebe). Brt. Guiana: Kartabo, Bartica Distr.; Kalacoon. Brazil:
Paraiba: Campina Grande (F. Pough); Independencia (W. M. Mann). Pernambuco:
Pernambuco. Minas Geraes: (Goldi, 1892; 0. P. Cambridge, 1902e); Diamantina
(E. Cohn); Corvelo (F. Pough); Bello Horizonte (D. Walters). Rio de Janeiro:
Rio de Janiero (Keyserling, 1891). Paraguay: (Simon, 1897k; F. P. Cambridge,
1902); Asuncion (E. Reimoser, C. J. D. Brown). Argentina: (Gerschman and
Schiapelli, 1943). Santiago del Estero: Termas de Rio Hondo, (M. L. Aczel). Chaco:
Resistencia (D. Freiberg).
Spain: (Galiano, 1910), probably a wrong determination.
Saudi Arabia: Dhahran (R. H. Daggy). Yemen: (Simon, 1890a). India:
(Pocock, 1900f; O. P. Cambridge, 1902e); Kolar (G. P. Staunton).
Libya: (Caporiacco, 1933) doubtful records. Sudan: (O. P. Cambrdige, 1902e;
Reimoser. 1928). Ethiopia: (Berland, 1922c); Schoa (Artinovi). Kenya: (Ber-
land, 1922c). Ruanda. Kenia: Naivashi 1900 met. (Berland, 1920a); Albert Lake
(Strand, 1913f). Zanzibar: (Karsch, 1878b). Mozambique: (Lessert, 1936); Lourengo
Marques (N. L. H. Krauss). Madagascar: (Strand, 1908e; 1916a; Hirst, 1911);
Tananarive (N. L. H. Krauss; V. J. Tipton); Ambositra. Southern Rhodesia. South
Africa: (Smithers, 1944); Port Shepstone (N. L. H. Krauss); Scottsville (R. F.
Lawrence); Maritzburg (R. F. Lawrence); Port Elizabeth (H. A. Spencer); Fendekuil,
Piquetberg Dist. Southwest Africa: (Simon, 1910c); Okakandje; Gobabis; Windhuk.
Congo: (Lessert, 1929; Giltay, 1935b); Poko. Cameroons: (Fage, 1927). Gold
Coast: Achimota, Acera (N. L. H. Krauss). Port Guinea: (Simon, 1907e). Sen-
egal: Dakar (E. H. Newcomb; N. L. H. Krauss).
Australia: (O. P. Cambridge, 1902e) a doubtful record; Darwin (B. Malkin).
Philippines: (Keegan, et. al. 1950). New Guinea: Hollandia (B. Malkin). New
Zealand: Finsch (Dahl collection), probably a wrong locality label. Hawaii: (Bianchi,
1944; Krauss, 1944).
MACTANS(FABRICIUS)
LATRODECTUS
Figures 1, 5-7, 19-21, 38, 53-55, 56-67, 72-79; Maps 2-5
Aranea mactans Fabricius, 1775, Systema Entomologiae, p. 432.
Female type from America (Lewin) has been lost.
Aranea tredecimguttatus Rossi, 1790, Fauna Etrusca, 2: 136, P1. 9,
Fig. 10, 9 [the date of the edition consulted was 1807 and the page 216].
Female type from Tuscany, Italy.
Latrodectus mactans, Walckenaer, 1805, Tableau des Araneides, p. 82.
Roewer, 1942, Katalog der Araneae, 1: 427 (in part). Bonnet, 1957,
Bibliographia Araneorum, 2: 2372 (in part).
Theridion lugubre Dufour, 1820, Ann. Gen. Sci. Phys., 4: 155, P1.
69, Fig. 1, 9. Type from Egypt.
Latrodectus erebus Audouin, 1825, Explication sommaires des planches
OFMAP
EXPLANATION
MAP 4. Distribution of dorsal abdominal setae of female Latrodectus mactans and
leg length. (Length of first patella and tibia divided by carapace length).
Small figures next to large ones indicate number of specimens whose measure-
ments were averaged.

rf
'...~
01r. ~~~?.
P t.-
iK^ \o :I
^^
-I-
WL
0)
as
0-1
JI
wH
..X 1. - vs
~
c
OD, E
o
o
0 E
j Ic

26 HERBERT W. LEVI
d'Arachnides in Savigny Descr. de l'Egypt, p. 137, P1. 3, Fig. 9, 9.

Female type from Salahyeh, Egypt.
Latrodectus argus Audouin, 1825, Ibid., Fig. 10, 9. Female type
from near Alexandria, Egypt.
Latrodectus venator Audouin, 1825, Ibid., p. 138, P1. 3, Fig. 11, 9.
Female types from near Alexandria, Egypt.
Meta hispida C. L. Koch, 1836, Die Arachniden, 3: 9, Fig. 166, 9.
Female type from vicinity of Nauplia [Nauplion], Greece.
Meta schuchii C. L. Koch, 1836, Ibid., 3: 10, Fig. 167, 9. Female
type from Greece.
Latrodectus malmignatus Walckenaer, 1837, Histoire Naturelles des
Insectes Apteres, 1: 642. New name for Araneus tredecimguttatus.
Latrodectusmartius Walckenaer, 1837, Ibid., 1: 137. Type from Italy.
Latrodectus oculatus Walckenaer, 1837, Ibid., 1: 645. New name for
Latrodectusargus.
Latrodectusformidablis Walckenaer, 1837, Ibid., 1: 646. Name given
to Fig. 191, p. 18 of Abbot's manuscript drawings which were made in
Georgia, United States.
Latrodectusperfidus Walckenaer, 1837, Ibid., 1: 647. Name given to
p. 18, Fig. 195 of Abbot's manuscript drawings which were made in
Georgia, United States.
Latrodectusvariolus Walckenaer, 1837, Ibid., 1: 648. Name given to
Fig. 194, p. 18, Fig. 391, p. 31, Fig. 396, p. 32 of Abbot's manuscript
drawings which were made in Georgia, United States.
Latrodectus intersector Walckenaer, 1837, Ibid., 1: 649. New name
for Araneus mactans varietas Fabricius.
Latrodectus quinqueguttatus Krynicki, 1837, Bull. Soc. Imp. Nat.
Moscou, 5: 75, P1. 6, Fig. 2, 9. Type from Cherson (?) and Odessa,
Russia.
Latrodectus conglobatus C. L. Koch, 1838, Die Arachniden, 4: 41,
Fig. 274, 9. Female type from Greece.
?Latrodectusdotatus C. L. Koch, 1838, Ibid., 8: 115, Fig. 683, c".
Male type from Pennsylvania, United States (may be L. curacaviensis).
Tetragnathazorilla Walckenaer, 1841, Histoire Naturelles des Insectes
Apteres, 2: 221, P1. 19, Fig. 2, ce. Male type from the Carolinas, United
States.
Theridion verecundum Hentz, 1850, Jour. Boston Soc. Nat. Hist.,
6: 280, P1. 10, Fig. 1, 2, c, 9. Types from Carolinas, Georgia, and
Alabama, lost.
Theridion lineatum Hentz, 1850, Ibid., 6: 281, P1. 10, Fig. 3, c".
Types from North Carolina and Alabama.
Theridion menavodi Vinson, 1863, Araneides des iles de la Reuion,
Maurice et Madagascar, p. 306, P1. 8, Fig. 5, c?. Type from Mada-
gascar. NEW SYNONYMY.
Latrodectuscinctus Blackwall, 1865, Ann. Mag. Nat. Hist., (3) 16: 341.
One female type from confluence of Shire and Zambesi Rivers,
Mozambique.
Latrodectushasselti Thorell, 1870, Oefers. Kongl. Vet. Ak. F6rhandl.,
27: 369. Juvenile female and male types from New Holland [Australia]
in the Riksmuseum Stockholm.
Latrodectus scelio Thorell, 1870, Ibid., 27: 370. Female type from
New Holland [Australia], in the Riksmuseum Stockholm.

Latrodectus katipo Powell, 1870, Trans. New Zealand Inst., 3: 57,

P1. 5, Fig. a-g, 9. Female types from New Zealand lost.
Theridion lineamentum McCook, 1879, Agricultural Ants of Texas, p.
203. New name for Theridion lineatum Hentz.
Theridium melanozantha Urquhart, 1886, Trans. New Zealand Inst.,
19: 102, P1. 8, Fig. 8, ci. Six male syntypes for Waiwera, New Zealand
in the Canterbury Museum.
Theridium zebrina Urquhart, 1889, Ibid., 22: 256. Female holotype
from Wellington, New Zealand in the Canterbury Museum.
Latrodectus katipo var. atritus Urquhart, 1889, Ibid., 22: 259. Type
locality from Portland Is., New Zealand.
Latrodectus scelio var. indica Simon, 1897, Bull. Mus. Hist. Nat.,
Paris, 3: 98. Female type from Mascate [? Oman, Arabia] in the Museum
National d'Histoire Naturelle in Paris.
Latrodectus elegans Thorell, 1898, Ann. Mus. Civ. Genova, 39: 293.
Three female types from Carin ChebA Mts., Burma in the Genova
Museum.
Latrodectus insularis insularis Dahl, 1902, Sitzungsber. Gesell.
Naturf. Freunde Berlin, p. 41. Female types from St. Thomas, Antilles
in the Berlin Museum.
Latrodectusinsularis lunulifer Dahl, 1902, Ibid., p. 21. Female types
from Haiti, Antilles in the Berlin Museum.
Latrodectus sagittifer Dahl, 1902, Ibid., p. 22. Female types from
Porto Alegro, Brazil, in the Berlin Museum. NEW SYNONYMY.
Latrodectusancorifer Dahl, 1902, Ibid., p. 43. Female types from New
Guinea in the Berlin Museum. NEW SYNONYMY.
Latrodectushahli Dahl, 1902, Ibid., p. 43. Female types from Bismarck
Archipelago and Phillippines in the Berlin Museum. NEW SYNONYMY.
Latrodectus luzonicus Dahl, 1902, Ibid., p. 44. Female types from
Philippines in the Berlin Museum. NEW SYNONYMY.
Latrodectus stuhlmanni Dahl, 1902, Ibid., p. 45. Female types from
"East Africa?", in the Berlin Museum. The specimens in the Berlin
Museum are marked as coming from Togo. NEW SYNONYMY.
Latrodectusrenivulvatus Dahl, 1902, Ibid., p. 45. Female types from
German Southwest Africa in the Berlin Museum. NEW SYNONYMY.
Latrodectus indistinctus 0. P. Cambridge, 1904, Ann. South African
Museum, 3: 154, P1. 11, Fig. 1, 9. Female types from Hamre Mission
Station, Malmesbury Division, Cape Peninsula, South Africa.
Latrodectus hasselti aruensis Strand, 1911, Abhandl. Senckenberg
Gesell., 34: 136. Female type from Aru Island, East Indies in the Sencken-
berg Museum, Frankfurt-am-Main. NEW SYNONYMY.
Latrodectusincertus Lawrence, 1927, Ann. South African Mus., 25: 30,
P1. 1, Fig. 19; P1. 4, Figs. 78, 79. Female syntypes from Ongandjera and
Namutoni, Southwest Africa in the South African Museum, Capetown.
NEW SYNONYMY.
Latrodectus albomaculatus Franganillo, 1930, Inst. nac. invest. cien.,
Habana, 1: 11. Female type from Cuba. NEW SYNONYMY.
Latrodectusagoyangyang Plantilla and Mabalay, 1935, Monthly Bull.
Bur. Health, Philippines, 15: 188. Type locality Cebu, Philippines.
Latrodectus mactans mactans, Chamberlin and Ivie, 1935, Bull. Univ.
Utah, Biol. Ser., 3(1): 13. Incorrectly stated type locality as Massachusetts
where this species does not occur.

28 HERBERT W. LEVI
Latrodectusmactans texanus Chamberlin and Ivie, 1935, Ibid., 3(1): 14,

Figs. 3, 15-18. Female holotype from Alvarado, Texas in the University
of Utah collection. NEW SYNONYMY.
Latrodectusmactans hesperus Chamberlin and Ivie, 1935, Ibid., 3(1): 15,
Figs. 1, 4, 6-14, 21, 23-33, c, 9. Female holotypes from Salt Lake
City, Utah in the University of Utah collection. NEW SYNONYMY.
Latrodectus indistinctus var. karooensis Smithers, 1944, Ann. South
African Mus., 36: 298, Fig. 16, 9. Female types from high lying parts
of the Karroo, South Africa probably in the South African Museum,
Capetown. NEW SYNONYMY.
LatrodectusrevivensisShulov, 1948, Ecology, 29: 209, Figs. 1-12, c?, 9.
Female and male (?) syntypes from 37 km. southwest of Beer Sheba near
the settlement of Revivim, Palestine probably in the collection of the
Hebrew University, Jerusalem. NEW SYNONYMY.
Latrodectus mactans mexicanus Gonzalez, 1954, An. Inst. Biol. Univ.
Mexico, 24: 455. No type specimens or type locality. Range: State
of Mexico to Chiapas. NEW SYNONYMY.
COMMENTS. Almost all authors from Hasselt (1860) to Smithers
(1944) have suggested that the species of Latrodectus they had studied,
and other species described in the literature, are probably synonyms of
L. tredecimguttatusor L. mactans. F. P. Cambridge, revising the genus
in 1902, distinguished only four species of the 43 described in the literature.
(L. hystrix, L. geometricus, L. pallidus, L. mactans) but was uncertain
about several other species (L. curacaviensis, L. menavodi, L. hasselti, L.
katipo) of which the males were not well known at the time. Gerschman
and Schiapelli (1943) considered L. hasselti, L. menavodi and L. katipo
to be subspecies or varieties of L. tredecimguttatus. Charitonov (1954) in
his discussion on Latrodectus referred to the six species of the genus.
Many other authors (e. g. O. P. Cambridge, 1904) have considered it
improbable that spider species could have a wide distribution and have
thus defended their species. The use of the names has been quite in-
consistent. For instance L. tredecimguttatushas been used for the form
occurring in Persia (Werner, 1936), Madagascar (Vinson, 1863; Lenz,
1891), East Africa to Mozambique (by many authors), to St. Helena
Island (0. P. Cambridge, 1873f, ex Bonnet, 1945); L. hasselti for specimens
from Arabia (Simon, 1897) to the Galapagos Islands (Marx, 1890a, ex
Bonnet, 1945); L. mactans from America to the Philippines (Sloggett, 1945)
and Hawaii (Krauss, 1944). Synonymizing these names is therefore not
an unprecedented action of mine, and additional specimens available in
collections at the present time strongly support F. P. Cambridge's ideas of
the genus.
The type locality of Latrodectus mactans is not known. Mrs. Ella
Zimsen, Copenhagen Museum (in. litt., 1958) wrote that Fabricius had
connections with an English person named Lewin. It is not likely that
this was John William Lewin, who collected in Australia, since this
naturalist was only five years old at the time of the description of L.
mactans. Probably his father, William Lewin, gave the specimens to
Fabricius. William Lewin was a naturalist and illustrator in London,
but is not known to have left England on collecting expeditions. Fabricius'
specimen has a red dorsal abdominal line; it may have come from the
southeastern United States or the West Indies. Koch's Meta schuchii

r\ 57 \ e58
(I
--
\61 61
n\ 63
64~r
67
f
FIGS. 56-67. Variation in Latrodectus mactans, left palpus. 56-58. (Colorado).

59-61. (Peru). 62-64. (Corsica). 65-67. (Tasmania). Notice tighter
embolus coil in individual palpus from Peru.

30 HERBERT W. LEVI
was considered to be L. tredecimguttatusby F. P. Cambridge, although

Simon 1881a, p. 179 (ex Bonnet, 1945) considered pallidus to be a synonym.
Cambridge has pointed out that all specimens labeled schuchii in collections
are black, spotless tredecimguttatus. Despite this, both Roewer (1942) and
Bonnet (1957) still consider it a separate species. Latrodectus menavodi
from Madagascar differs from African specimens only in abdominal
spination. Many specimens were available from Madagascar. Although
Latrodectus cinctus is a much older name than indistinctus, the older
name rarely has been used. Specimens were available from near its type
locality in Mozambique. The specimens on which Dahl's species were
based were available for study, as was the type of Latrodectus hasselti
aruensis Strand. No determined specimens of L. incertus were available;
however, many specimens from Southwest Africa were examined. Dr.
Lawrence, its author, wrote in a letter, 1957, "I am pretty certain it [L.
incertus] is only an immature indistinctus, certainly not a separate species
although possibly a subspecies."
Problems concerning Chamberlin and Ivie's subspecies based on
coloration and size have been mentioned above. Latrodectusalbomaculatus
of Cuba was considered by its author, in a later paper (1936), to be a
subspecies of mactans. Latrodectusm. karooensis is a color variety found
in various parts of South Africa, probably not a geographic subspecies.
No significant differences were found between L. revivensis and sympatric
specimens determined as L. tredecimguttatus,although its author, Shulov,
(in. litt. 1938) writes that "it is definitely and specifically different from
the L. XIII-guttatus. The males as well as the young spiders are of
completely different colour and design. The web is different and so is
its placement regarding the shrubs." Latrodectus mactans mexicanus
was described to provide a name for the variously colored specimens found
in Mexico.
DIAGNOSIS. The epigynum of L. mactans is similar to that of L.
curacaviensis; however, mactans females differ from those of the latter
species as well as from L. hystrix in having four or five loops in dorsal
view of the internal genitalia (Figs. 19-21), while L. curacaviensis and
L. hystrix have three or less. The four internal coils are wider than those
of L. geometricus,also the lower lip of the opening of the epigynum of L.
geometricus projects towards each side, while it usually does not in L.
mactans.
The embolus of the male palpus of L. mactans has three coils, (Figs.
5-7) while that of L. curacaviensis has only two, and L. geometricushas
four.
VARIATION. The color, size, leg-length vary geographically; ia
America paralleling L. curacaviensis (Maps 3-5). The female ducts are
somewhat shorter in South African specimens, and so may be the palpal
embolus; however this is the only geographic variation observed in diag-
nostic characters.
Other characters examined show that there is no relationship in the
variation of individual characters, and that dines go in various directions
(Maps 3-5). Females with the lightest coloration come from South
America and Southwest Africa. The "hour glass" on the venter of the
female has been mapped only for North American specimens (Map 5).
It is present in specimens from various parts of the world, including

carapace length above line

leg length below line
L. mactans
pattern on abdomen
left dorsum
right venter
MAP 5. Latrodectus mactans. Upper map: Distribution of female carapace length

in mm., above line, female leg length (length of first patella and tibia divided
by carapace length) below the line. Plus sign next to line indicates average
based on five or more specimens, one next to line indicates that the measure-
ments are those of an individual. Lower map: Distribution of dorsal (left)
and ventral (right) abdominal patterns of females.

32 HERBERT W. LEVI
New Guinea and New Zealand. It is r-shaped in Panama; almost square

in specimens from Veracruz, Mexico, Peru, and Tasmania; consists of
only two spots in specimens from California and Western Australia.
Many specimens from South Africa, including the light specimen from
Lake Nyasa, have the venter black without any spots. The largest-sized
females come from Mediterranean region and California; the smallest
from South America and Northern Australia. Those with the longest
legs are found in southern Mexico and New Guinea, those with the shortest
legs, in New Zealand and the Canary Islands. The females with the
strongest abdominal spination are from the Mediterranean region, those
with the weakest spination are from America, although females lacking
strong spines are found in Madagascar and New Zealand.
Abdominal spines vary most within Europe (Map 4). The strongest
spines were found in specimens from Palestine. The short, thick spines
of Corsican specimens have an additional basal spine, while those of the
neighboring Sardinia (not illustrated) are relatively weak and lack the
basal spine. Abdominal spines are very similar in specimens from dif-
ferent parts of America, where color variation is greater (Map 3) than in
other regions. The most uniform coloration is found in specimens from
the Australian region; however there is some variation in this region in
abdominal spination, and extreme variation in leg length. It should be
kept in mind that there is often considerable variation in spiders from the
same locality. The carapace length of nine females from Bulawayo,
Southern Rhodesia, varied between 4.8 and 5.8 mm.; the leg length
(length of first patella-tibia carapace length) from 1.62-1.76.
As mentioned above, there are many other characters, including the
extreme variation in carapace shape, that were not studied at all. Phy-
siological characters had to be ignored, although variation in potency of
venom or number of bites has been reported (Crome, 1956; O'Rourke,
1956). On the other hand, this variation may reflect differences of the
spiders' habitat or abundance, or possibly of Man's habits. In a recent
paper, Maretic (1951) reported that no spider bites occurred in Istria
before 1948; however, between 1948 and 1951 there were 26, perhaps
indicating fluctuation in numbers of spiders.
Of all material studied, however, specimens from North America and
Europe differ most, whereas specimens from Peru and the Australian
regions are virtually indistinguishable.
It is of interest that there are no references to poisonous spiders in the
Bible (Fr. Chrysanthus pers. comm.). Perhaps the range of the species
was discontinuous for a period of time.
"Subspecies" have been used by different authors, particularly in
Europe and Africa, to indicate local variation of color and spines. These
trinomials were derived from older names (lugubris, erebus,schuchii). The
use of subspecific names for regional character complexes de-emphasizes
the variability of these characters, and would discourage study of the
many traits necessarily ignored here. O'Rourke (1956), for instance,
suggested basing subspecies on relative toxicity of venoms, since no
fatalities are reported from Canada. He overlooked several other pos-
sibilities: that the number of fatalities may reflect the size of the spider
or the abundance of human outhouses (a likely place for bites to occur);
and that he was dealing with misidentified specimens.

The study of populations within the species should emphasize the

distribution of the variable characters rather than the artificial grouping
of these characters into geographical entities: subspecies. To follow the
latter course would increase nomenclature confusion, and furthermore,
there is no limit to the number of names that could be made. To maintain
reference with older usage, however, it might be advisable to recognize
for the time being the following subspecies:5
Latrodectusmactans mactans, Americas, rarely having strong spines on
the abdomen.
Latrodectus mactans tredecimguttatus, Europe and Mediterranean
region, abdomen covered by many strong short and long spines; sometimes
dorsum with series of red spots; short, strong legs.
Latrodectusmactans cinctus, southern portions of Africa, with weaker
spines than tredecimguttatus,smaller size, and usually longer legs.
Latrodectus mactans menavodi, Madagascar, Seychelles, abdominal
spines uniform in size.
Latrodectusmactans hasselti, India to New Zealand, abdominal spines,
leg length variable; abdomen usually with a posterior dorsal median
white line.
It should be pointed out again that this is an artificial classification;
that the only one of these subspecies which is distinct in more than one
character examined (although the characters are not of diagnostic im-
portance) is L. mactans tredecimguttatus,whose characters grade into the
neighboring forms. The distribution pattern of characters is too irregular
to permit delineation of subspecies from smaller geographic areas (see
Map 5).
HABITAT. In the southern United States this species is found "under
stones and pieces of wood on the ground, about stumps, in holes in the
ground and about outbuildings" (Comstock, 1912). In South Africa the
web "is constructed at the base of small bushes, tufts of grass or stubble
or among heaps of loose debris . . .Nests have been found under the
overhanging edges of stones and in rock crevices as well as among occa-
sional rubbish such as old tins, but sites as these are the exception. No
nest . . . has yet been recorded from the immediate precincts of a build-
ing" (Smithers, 1944). In Palestine this species is found "in strong
dusty webs under large stones, as well as in the holes of rodents, near the
roots of shrubs and, rarely, in the corners of old buildings and in the
entrance of caves. The webs are never found among shrubs, as in the
case of other countries." (Shulov, 1940). [This spider] "lives in the
desert and semidesert region and also in some sections of the steppe, in
areas close to semidesert" in Russia (Charitonov, 1954). In New Zealand
the spider is found only high on ocean beaches and river beds, under
driftwood, trash and in the base of vegetation (B. Holloway, R. Forster,
pers. comm.). The spider's habitat is similarly limited to beaches in
Jamaica, British West Indies (A. M. Chickering, pers. comm.).
DISTRIBUTION. In warm areas of all continents.
RECORDS. United States. New York: Bronx; Brooklyn (Spector); Queens (Mum-
ford); Nassau Co.: Valley Stream (L. Wiley). New Jersey: Burlington Co.: Rancocas
5F. Nikoli6 (in lett. 1958) indicated that he independently revised Latrodectus
and considers tredecimguttatus and indistinctus subspecies, menarodi, hasselti and
Katipo varieties. His paper is in press in a Yugoslavian journal.

34 HERBERT W. LEVI
Pk. (E. A. Richmond). Ocean Co.: sev. coll. Pennsylvania: Fayette Co. (sev.
coll.). Delaware: Kent Co.: (R. Traub). Maryland: Harford Co.: (J. Davis);
Washington Co.: (J. Hyslop). Virginia: Shenandoah Co.; York Co. West Vir-
ginia: Ohio Co.: Wheeling (K. Haller); Pocahontas Co. (K. Haller). North
Carolina: Buncombe, Burke, Carteret, Madison, Perquimans, Swain, Wake Counties.
South Carolina: Charleston, Florence, Horry Counties. Georgia: Charleton, Cha-
tam, Cobb, Dade, Floyd, Mitchell, Rabun, Richmond Counties. Florida: Alachua,
Charlotte, De Soto, Escambia, Gilchrist, Hillsborough, Jackson, Jefferson Lake,
Liberty, Levy, Monroe, Okaloosa, Orange, Osceola, Polk, Pinellas, Putnum, Volusia,
Walton Counties. Alabama: Baldwin, Green, Houston, Lee, Sumter, Tallapoosa
Counties. Mississippi: Covington, Forrest, George, Simpson, Yazoo Counties.
Louisiana: Beauregard, Calcasieu, Caddo, East Baton Rouge, Jefferson, Davis,
Lincoln, Madison, Orleans, Saint Landry. Saint Tammany Parishes. Ohio: Athens
Co.: Lodi (W. C. Stehn); Brown Co.: Georgetown (G. Noland); Clinton Co.:
Martinsville (W. Wilson). Indiana: Clark Co.: Charlestown (H. Spieth); Craw-
ford Co.: Wyandotte (N. Banks); Martin Co.: Shoals (N. Banks). Kentucky:
Boyle, Breathitt, Edmonson, Hart, Madison Counties. Tennessee: Knox, Sevier
Counties. Illinois: Alexander, Cumberland, Effingham, Jackson, Jefferson, Ma-
coupin, Madison, Monroe, Pike, Pope, Wabash, Wayne, Williamson Counties.
Missouri: Greene, Iron, Phelps, St. Louis, St. Francois Counties. Arkansas: Ben-
ton, Izard, Hempstead, Washington Counties. Kansas. Harper Co.: Harper.
Oklahoma: Grady, Muskogee, Payne Counties. Texas: most central, southern
and eastern counties. Colorado: Boulder, Denver, Douglas, Eagle, El Paso, Fre-
mont, Huerfano, Jefferson, Larimer, Mesa, Prowers Counties; Mesa Verde Natl.
Park. New Mexico: Dona Ana, Eddy, Grant, Hidalgo, Lincoln, Otero, San Miguel
Counties. Arizona: Conchise, Coconino, Gila, Graham, Mohave, Maricopa, Navajo,
Santa Cruz, Yuma Counties; Organ Pipe Cactus Natl. Mon. Utah: Millard Co.
Salt Lake Co.: Dry Canyon (W. J. Gertsch); Logan Canyon (G. Knowlton); Mill
Creek. Washington Co. Nevada: Las Vegas. Oregon: Eugene (B. Malkin).
California: Alameda, Contra Costa, Fresno, Imperial, Inyo, Los Angeles, Marin,
Napa, Placer, Riverside, San Benito, San Diego, San Francisco, San Luis Obispo,
Santa Barbara, Santa Clara, Santa Cruz, Solano, Tulare, Yolo Counties; Sequoia
Natl. Pk.
Mexico. Tamaulipas: Nueva Morelos; Villa Juarez; Jim6nez; Abasolo; Santa
Teresa; Cuidad Victoria; San Jose. Nuevo Leon: Monterrey; Linares. Coahuila:
72 mi. E. of San Pedro; Saltillo; Guadalupe; Nava; Sierra de la Encontada. Chihuahua:
Villa Ahumada; Santa Barbara; Madera; Primavera; La Sauceda, 7000 ft.; Las Canoas
Babicora; Sierra de en Medio, 5000-5700 ft.; Gran Morelos; Huejotitlan; W. of
Camargo; 22 mi. N. of Parral. Sonora: Desemboque; Tastiota; Yavaros; Guaymas;
Alamos; Isla Pelicano; 20 mi. SW of Sonoita. Baja California: Isla Santa Ines,
Gulf of California; N. of Ensenada; Tia Juana; Palmarito de Abajo; Isla Raya; Patos
Isl.; Isla Partida; Santa In6s Is.; San Francisquito Bay. San Luis Potosi: Tepetate;
nr. Cueva de los Savinos nr. Valles; Huichichuyan. Durango: Santa Maria del
Oro; 26 mi. S. of Loredo; Providencia; Durango; Las Puentes; Ojo de los Encinos;
Otinapa; San Isidro; 60 mi. NW of Durango; Nombre de Dios; Palos Colorados,
8000 ft.; El Salto; 40 mi. NE of El Salto. Sinaloa: nr. Culiacan. Nayarit: Tepic;
Compostela; Jesis Maria; Arroyo Canaveral; Jalisco; Rosamorado. Veracruz:
Perote; Tecolutla; Franca Vieja; Cordoba; Papantla; Boca del Rio; Tampico. Hidalgo:
Ixmiquilpan; Guerrero Mill. Puebla: Acatlan; Tehuacan; Tlacotepec; Teziutlan.
Morelos: Cuernavaca; Cuajomulco; Alpoyeca; Cocoyoc; Temixco. Distrito Federal:
common. Mexico: Tenancingo; Ixtapan; Highway 71, km. 125. Jalisco: Atenqui-
que; Chapala; Ajijic; Barranca de Oblatos Guadalajara; Tlaquepaque; Mazamitla.
Michoacdn: Tzararacua Falls, nr. Uruapan; Mirador de Atzimbo; Lk. Patzcuaro.
EXPLANATION OF MAP
MAP 6. Latrodectus curacaviensis. Left map: Specimens examined. Center map:

Distribution of female leg length (length of first patella and tibia divided
by carapace length). Right map: Distribution of female dorsal abdominal
pattern and distribution of carapace length in mm. Small figures next to
larger ones indicate number of specimens whose measurements were
averaged.

SPIDER GENUS LA TRODECTUS 35
(0
c
0
0
o-
J

36 HERBERT W. LEVI
Colima: common. Guerrero: Mexcala; 39 mi. S. of Chilpancingo; Chilpancingo;

Acapulco; Taxco; Iguala; 38 mi. S. of Iguala; Teloloapan; 62 mi. N of Acapulco.
Oaxaca: San Ber6nimo; Tehuantepec; Salina Cruz; San Mateo del Mar; Palomares.
Chiapas: Arriaga S. of Arriaga Mts.; Las Cruces; Puerto Madero; San Carlos;
Tuxtla Gutierrez. Campeche: San Dimas.
Guatemala: Yepocapa, 4300 ft. (T. H. Farr). Honduras: Tegucigalpa (Clarke).
Nicaragua: San Marcos (C. F. Baker). Panama: La Campana (E. Fitcher);
Chorrera, Panama Prov. (C. and M. Michener).
West Indies. Bahiama Is.: Mangrove Cay; New Providence; Cat Is.; Hog Cay;
Long Is. Cuba: Habana; Marianos; N. of Vinales; Santiago de las Vegas; Lomas
de Comoa, Habana; Soledad; Antonio de Bafios; Cienfuegos, San Jos6; Siboney; San
Vincente; P. del R.; Cojimar; Bolondr6n; Jiguani; coast nr. Pico Turquino; Sierra de
Rangel, P. del R.; Maisi. Isla de Pifios. Carribean Is.:Cayman Brac Is.; Navassa
Is. Jamaica: St. Andrew, St. Ann, St. Catherine, St. James, Portland parishes.
Haiti: Port au Prince; St. Marc; Kenscoff; Petionville; Grand Riviere; Cap Haitien.
Dominican Rep: Barahona Harbor; San Domingo City; nr. La Romana; San Jos6
de las Matas; Mt. Diego de Ocampa; Barahona; Puerto Plata. Puerto Rico: Mona
Isl.; Ponce; Las Mesas; Maricao; Aguirre; Rio Piedras. Vieques Isl.; St. Thomas Isl.
Venezuela: Maracay. Colombia: Popayan, Cauca; 5 km. N. of Villavieja,
Huila. Ecuador. Sta. Elena (C. M. Breder). Peru: Oroya; Ninamba, Rio Pampas.
Abancay: 2460m. Arequipa: Mollendo, Loma Zona; Arequipa. Camana: 12 mi.
SE of Camana. Cuzco: Ollantaitambo; Cuzco; Marcapata Riv. Huancavelica:
Mejorado, Rio Mantaro 3200 mi. Junin: Huancayo. Lima: 16 mi. NW of Chancay.
Oriente: Barranca. Otuzco: Otuzco, 2600 m; La Libertad. Tarma: Tarma,
3100 m. Brazil: Pernambuco: Recife. Rio Grande do Sul: Porto Alegre (F.
Dahl coll.). Paraguay. Chaco: Nanahua (G. S. Carter).
Portugal: (Bacelar, 1928; 1933c). Spain: (Galiano 1910). France: (0. P.
Cambridge, 1902e; Berland, 1930d; Denis, 1931a). Corsica. Sardinia: Sorgono
(E. Reimoser). Italy: (Reimoser, 1920a); Prov. Grosseto, Montepescali (R. Dow);
Busento Valley. Yugoslavia: (Chvzer and Kulczynski, 1894; Reimoser, 1920a;
Kolosvary, 1936b). Greece: Tripolis (Cecconi, 1895; Fage, 1921b). Russia:
(Charitonov, 1932b); Don.
Turkey: (Kulczynski, 1903a); Taurus Mts., Buroock. Syria: (O. P. Cam-
bridge, 1872a). Palestine: Jerusalem; Beer-Sheva (A. Shulov). Arabia: (Pocock,
1895c). India: (Simon, 1902d; Reimoser, 1934a; Dyal, 1935; Keegan, 1955); Kotar.
Nepal: Mayangdi Khola nr. Darban, 28?24' :83023', 3000 ft. (K. H. Hyatt). Indo-
china: (Simon, 1908a). Malayan Archipelago: (Hasselt, 1890e; Reimoser, 1925).
Canary Isl.: (O. P. Cambridge, 1902e); Gratiosa. Morocco: (Simon, 1909d).
Tunisia: (Strand, 1908). Egypt: numerous references referring to "Egypt".
Sudan: (Simon, 1906e, questionable record). Ethiopia: (Strand, 1908i; Berland,
1922c, questionable records, may have been confused with L. geometricus). Uganda:
Lessert, 1915a; Berland, 1920a, questionable record). Kenya: (Berland, 1920a,
questionable record; Caporiacco, 1949). Mozambique: Lk. Nyasa. Seychelles:
Farquhar (J. S. Gerdiner). Madagascar: (Strand, 1908d); Jananarivo; Mahanoro;
Ankasoabo (B. Ljungquist). South. Rhodesia: Bulawayo; Plumtree; 44 mi. on Falls
Rd.; Kumalo (P. Pringle). South Africa: (Smithers, 1944); Touthansberg; Transvaal;
Pietermaritzburg (R. F. Lawrence); Capetown; Jansenville; S. Lucia Lk.; Grahams-
town; Steinkopf (Dahl coll.). Southwest Africa: Okahandja; Windhuk; Kubub-
Monz; Kamagams (Dahl coll.) St. Helena: (O. P. Cambridge, 1873f). Togo:
(Dahl coll.), questionable locality. Senegal: (F. P. Cambridge, 1902e); Dakar
(E. H. Newcomb).
Philippines: (Sivickis 1928; Asis 1934; Keegan et al. 1950; 1952); Luzon (Dahl
coll.); Cebu (Dahl coll.). Celebes: (Merian, 1911; Strand, 1915b). Halmahera:
(Sloggett, 1946). Timor: (Sloggett, 1946). New Guinea: (Rainbow, 1899a);
Hollandia (many coll.); Astrolabe; Port Moresby. New Britain: Herbertshbhe;
Ralum. New Ireland: (Strand, 1915b). Trobriand Is.: (R. L. Horton). New
Caledonia: (Berland, 1924); 7 mi. SE of La Foa (C. Remington). Australia: (L.
Koch, 1872d; Bradley, 1877; Frost, 1891a; Hobb, 1896; Simon, 1896k; Tidswell, 1907;
Simon, 1908e; Rainbow, 1912b; 1916a; Froggatt, 1920; Barrett, 1936). Western
Australia: Perth; Sir Graham Moore Is. (B. Malkin); Beverly (W. A. Butler);
Tammin (B. Y. Main). Northern Territory: 70 mi. S. of Darwin (B. Malkin);
Darwin (B. Malkin). Queensland: Townsville (B. Malkin); nr. Brisbane (G.
Found). Canberra (E. Britten). South Australia: Adelaide (E. Britten). Tas-
mania: Risdon (V. V. Hickman). New Zealand: (Butler, 1870; Fyffe, 1901; F. P.

68 69 70 71
72
72 73 74 I
75 _
I 1.
1,% I
\ I1-
:I Y
76 ,, 77 78 79
I /-\,m
\I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1
1
so 21 81
FIGS. 68-71. Variation in female genitalia of Latrodectus curacaviensis. 68. (Para-

guay). 69. (Cordoba, Argentina). 70. (Florida). 71. (Oregon).
FIGS. 72-79. Variation in female genitalia of L. mactans. 72. (Arkansas). 73.
(Virginia). 74. (Florida). 75. (Peru). 76. (France). 77. (Israel).
78. (South. Rhodesia). 79. (Tasmania).
FIGS. 80-83. Variation in female genitalia of L. geometricus. 80. (Florida). 81.
(Cuba). 82. (Saudi Arabia). 83. (Darwin, Australia).

38 HERBERT W. LEVI
Cambridge, 1902e); Hawkes Bay (T. Lomas); Christchurch (Howe); Waitere Beach,
N. of Wellington. Mariana Isl.: (Keegan, 1955). Hawaii: (Bianchi, 1944; Krauss,
1944); Honolulu (W. T. Doren).
LATRODECTUS PALLIDUS O. P. CAMBRIDGE

Figures 22-24, 31, 33-34; Map 7
Latrodectuspallidus 0. P. Cambridge, 1872, Proc. Zool. Soc. London,
p. 287. Females from the Plains of Jordan, Palestine. Roewer, 1942,
Katalog der Araneae, 1: 425. Bonnet, 1957, Bibliographia Araneorum,
2: 2377.
?Latrodectuspallidus immaculatus Caporiacco, 1933, Ann. Mus. Civ.
Genova, 56: 323. Type from Gialo Oasis Kufra in Libya.
Latrodectuspallidus pavlovskii Charitonov, 1954, [Zool. Journ.], 33: 480,
Fig. 1, 2, cd. Male type from village of Kara-Bogez, 40 km. north of
Kizil-Arvat, Turkmen, Russia. NEW SYNONYMY.
COMMENTS. There is some doubt that Latrodectuspallidus immaculatus
is L. pallidus; it could be L. geometricus. Caporiacco's subspecies differed
in that the specimen he examined had slightly different coloration.
Charitonov's subspecies differed in "being slightly smaller, the lack of a
large number of supplementary dark spots (besides the four main ones)
on the dorsal side of the abdomen", the coloration of the venter, the
shape of the aperture of the epigynum, and in differences in hairs.
DIAGNOSIS. Only several female specimens were available. They
differed markedly in coloration from L. mactans. The abdomen looks
leathery (Fig. 31), possibly due to the white coloration. The epigynum
and internal genitalia are as in L. mactans.
Although males have been collected both in Palestine and Turkmen,
none could be obtained for this study. Charitonov's figures suggest that
the palpus is similar to that of L. mactans.
HABITS. The only evidence, except for coloration, that this is a
separate species from L. mactans, is reported differences in habits and
physiology. In Palestine L. pallidus is limited to the plains, while L.
mactans is widespread. L. mactans lives on the ground, while L. pallidus
lives in shrubs and feeds on ants, Messor semirufus (Shulov, 1940). On
the other hand, L. revivensis, believed to be a synonym of L. mactans,
also is reported from webs above ground (Shulov, 1948). Oviposition of
L. pallidus occurs in late summer and fall; L. mactans produces eggs in
winter. The thermal death point is higher (50.5 degrees C.) than that
of L. mactans (47.1 degrees C.). The venom is believed to be weaker.
In Turkmen the spider has been found in a river bed (Charitonov, 1954).
RECORDS. Russia: (Charitonov, 1954; at Kara-Bogaz, 50 km. N. of Kizil-Arvat,
Charitonov, 1955). Syria: (Schulov, 1940). Palestine: (O. P. Cambridge, 1872),
Beer-Sheva (Schulov). Iran: Bushire. Egypt. (Schulov, 1940). Libya: (Caporiacco,
1933), a doubtful record. Tripolitania: (Shulov, 1940).
LATRODECTUS CURACAVIENSIS (MULLER)
Figures 2-4, 16-18, 35-36, 38, 68-71; Map 6
Aranea curacaviensis Muiller, 1776, Des Ritters Carl von Linne voll-
standigen Natursystems, Suppl. Vol., p. 242. Female type from CuraCao
in the Lesser Antilles.
LatrodectusvariegatusNicolet, 1849, in Gay, Historia de Chile, Zoologia,
3: 461, P1. 4, Fig. 9. Female from Chile. NEW SYNONYMY.

Latrodectusthoracicus Nicolet, 1849, Ibid., 3: 462. Female type from

the Province Concepcion, Chile. NEW SYNONYMY.
Theridium mirabile Holmberg, 1876, An. Agr., Rep. Argentina, 4: 34.
Female from southwest of the village of Mercedes de Patagones,
Argentina. NEW SYNONYMY.
Latrodectus curacaviensis, Thorell, 1873, Remarks on Synonyms of
European Spiders, p. 511. Roewer, 1942, Katalog der Araneae, 1: 427.
Latrodectus malmignatus var. tropica Hasselt, 1860, Tijdschr. Ent.,
3: 62, P1. 5, Figs. 1-6, cd. Female type from Curacao, Lesser Antilles.
NEW SYNONYMY.
Latrodectus carolinum Butler, 1877, Proc. Zool. Soc. London, p. 75,
P1. 13, Fig. 3. Female syntypes from Charles Island, Galapagos. NEW
SYNONYMY.
Latrodectusapicalis Butler, 1877, Ibid., p. 75, P1. 13, Fig. 5. Female
syntypes from Charles Island, Galapagos. NEW SYNONYMY.
Latrodectusgeographicus Hasselt, 1888, Tijdschr. Ent., 31: 176, P1. 5,
Figs. 1-5, 9. Female type from Dutch Guiana.
Chacoca antherata Badcock, 1932, Jour. Linnean Soc. London, 38: 10,
Fig. 5. Male type from Nanahua, Paraguayan Chaco in the British
Museum, Natural History. NEW SYNONYMY.
Chacoca flavodorsata Badcock, 1932, Ibid., 38: 11, Fig. 6. Female
type from Nanahua, Paraguayan Chaco lost. NEW SYNONYMY.
Chacoca distincta Badcock, 1932, Ibid., 38: 11, Fig. 7. Juvenile female
type from Bauri, Paraguayan Chaco in the British Museum, Natural
History. NEW SYNONYMY.
Latrodectus carteri Badcock, 1932, Ibid., 38: 12, Fig. 8. Female type
from Nanahua, Paraguayan Chaco in the British Museum, Natural
History. NEW SYNONYMY.
Latrodectus cretaceus Badcock, 1932, Ibid., 38: 12, Fig. 8. Female
type from Makthlawaiya, Paraguayan Chaco in the British Museum,
Natural History. NEW SYNONYMY.
Latrodectus mactans var. bishopi Kaston, 1938, Florida Ent., 21: 60.
Male holotype from Lake Worth, Florida in the American Museum of
Natural History. NEW SYNONYMY.
Latrodectusfoliatus Mello-Leitao, 1940, Rev. Museo La Plata, 2: 35,
Fig. 4, cd. Male holotype from Buenos Aires, Argentina, in the Museo de
la Plata. NEW SYNONYMY.
Latrodectus curassaviensis, Bonnet, 1957, Bibliographia Araneorum,
2: 2367.
COMMENTS. The only Latrodectus found in collections from the
Dutch Lesser Antilles is this species of which specimens have been examined
from Aruba. Miiller indicated in the description that the species never
is found in cities, but is found among vegetation in the country, sometimes
in country houses, and that its bite is poisonous. Latrodectusnames with
type localities within the range of curacaviensis, and outside the range of
mactans, have here been synonymized arbitrarily with curacaviensis (L.
variegatus, L. thoracicus, L. mirabile, L. geographicus, L. malmignatus
tropica, L. carolinum, L. apicalis), although some previously have been
synonymized with mactans. A male palpus of Chacoca antheratahas been
examined and was found to be L. curacaviensis. The female type of C.
flavodorsatahas been lost, but since it was collected with the male of C.

40 HERBERT W. LEVI
antherata we can assume it to be L. curacaviensis also. The type of C.

indistincta is immature; it may be either L. geometricusor L. curacaviensis,
and my synonymy here is arbitrary. A paratype of L. carteriwas found
to be L. curacaviensis. Latrodectus cretaceous, according to Badcock,
differs only in coloration from L. carteri; it is undoubtedly the same
species. One female determined by Miss de Pikelin and Prof. Schiapelli
as L. foliatus was found to be L. curacaviensis,although the ducts are longer
than usual in Argentina and Paraguay specimens, and the coloration also
is different. Mello-Leitao's figure indicates that the palpus is that of
L. curacaviensis.
DIAGNOSIS. I have been unable to find differences in the epigynum
(Figs. 35, 36) between L. mactans and L. curacaviensis. In the female
genitalia, the connecting ducts of L. curacaviensis are shorter, having only
three loops in dorsal view. In the United States, the average length of
the first patella and tibia is greater than in L. mactans (Fig. 38) wherever
the two species occur together. Reese (1940, Jour. Hered., 31: 118)
described an "interesting" variation in female specimens of "L. mactans".
A fifth of his largest spiders had no marks on the dorsum of the abdomen,
while only a thirteenth of the smaller specimens were unspotted. Probably
Reese's smaller specimens [from West Virginia?] were L. curacaviensis, his
larger one L. mactans.
The palpal embolus is shorter and wider than that of L. mactans
(Figs. 2-4) in the United States. The males often are larger in size than
those of L. mactans.
VARIATION.The largest male came from Florida, the smallest from
Argentina. Dorsal abdominal color patterns of females are illustrated by
Map 6. The largest females came from the Pacific coast states of the
United States, the smallest from Argentina (Map 6). The females with
the longest legs came from Florida, those with the shortest legs from
Argentina, Galapagos Islands, and the Pacific United States. The ducts
of females from Paraguay and Argentina are shorter (Fig. 68), much like
those of L. hystrix, however, a female from Buenos Aires had the ducts
longer and the coloration different from other Argentinian specimens.
It was identified by Miss de Pikelin and Miss Schiapelli as L. foliatus.
This specimen may have been introduced from another part of the range.
HABITAT. L. curacaviensisis reported from trees and shrubs in Florida,
while L. mactans lives on the ground. In the northern states, L.
curacaviensis is found in fields and woods underneath logs and stones,
usually away from human habitations.
DISTRIBUTION. Americas from southern Canada to Patagonia, but
apparently absent in Mexico, Central America and Greater Antilles, and
more common in the temperate regions of North and South America.
The greater number of curacaviensis specimens from northern, compared
to southern, United States, may be due to the comparative rarity of
black-widows in the northern states. The extreme commoness of L.
mactans in the southern states probably discourages collectors from picking
up black-widows.
RECORDS. Canada. Ontario: Bruce Co.: Dyer Bay (D. B. Pengelly); Lambton
Co.: nr. Oaksdale. Norfolk Co.: Delhi (J. W. Watts). Alberta: Medicine Hat
(P. L. Pender). British Columbia: Kamloops (many coll.); Casade (B. Malkin);
Wellington (R. Guppy); Vernon (H. Leech); Vancouver (N. Banks); Victoria (many
coll.); Summerland (W. L. Gartrell).

United States. New Hampshire: Shirley Hill (J. H. Emerton). Belknap Co.:
N. of Sanborton. Cheshire Co.: Keene (A. Sheldon); Dublin (J. H. Emerton).
Grafton Co.: Plymouth (A. Gray). Vermont: Windham Co.: S. of Newfane (E.
B. Bryant). Massachusetts: Barnstable Co.: Woodshole. Martha's Vineyard
(W. Black). Middlesex Co.: Concord (Seton; W. Schevill); Tyngsboro (J. H.
Emerton). Norfolk Co.: Sharon (E. B. Bryant); Dover; Blue Hills (J. H. Emerton).
Connecticut: Litchfield, Middlesex, New Haven, New London Counties. New York:
Kings, Nassau, Suffolk Counties. Maryland: Montgomery, Prince Georges Counties,
Virginia: Fairfax, Northampton Counties. North Carolina: Raleigh. South Caro-
lina: Charleston. Florida: Highlands, Liberty, Marion, Orange, Palm Beach,
Putnam Counties. Louisiana: "Chastine" (K. P. Schmidt). Michigan: Benzie
Co.: Crystal Lk. (J. Hopkins). Berrien Co.: Lakeside (D. C. Lowrie). Livingston
Co.: George Reserve (H. K. Wallace). Wexford Co.: Cadillac (K. Sikkema).
Indiana: Porter, Putnum Counties. Kentucky: Breathitt Co. Tennessee: Roane Co.
J .41
:. \ . /-.. . . ....
* L.
*? L. pallidus
pollidus "
? L. dahli .
*a L.
L. hystrix
hystrix \-// /
MAP 7. Distribution of Latrodectus pallidus, L. hystrix, and L. dahli.
Wisconsin. Crawford Co.: Gays Mills (L. Kegel). Sauk Co.: Baxter's Hollow (H. and
L. Levi). VernonCo.: Wildcat Mtn. (H. and L. Levi). Illinois: Coles, Effingham,
Kankakee, La Salle, Lawrence, Pope Counties. Missouri: Green, Phelps Counties.
South Dakota: Pennington Co.: (V. Roth). Nebraska: Badlands. Kansas: Meade
Co. Texas. Brewster Co.: Chisos Mts. Basin (W. J. Gertsch and M. Cazier).
Pecos Co.: 15 mi. E. Ft. Stockton (W. J. Gertsch). Starr Co.: Rio Grande.
Terrell Co.: Dryden (C. Michener). Webb Co.: Laredo (R. Baird). Montana:
Custer, Daniels, Flathead, Gallatin, Ravalli Counties. Wyoming: Sheridan,
Weston Counties. Colorado: Boulder, Larimer, Mesa Counties. New Mexico:
Bernalillo, Luna Counties. Idaho: Gem, Nez Perce Counties. Utah: Salt Lake,
Tooele, Weber Counties. Washington: Benton, Columbia, Grant, Pierce, San Juan,
Yakima, Walla Walla Counties. Oregon: Benton, Jackson, Jefferson, Malheur,
Umatilla, Wallowa, Yamhill Counties. California: Humboldt Co.: Ft. Seward
(H. Essig). Marin Co.: Taylorville (H. Dietrich). Siskiyou Co.: Dorris (W. J.
Gertsch).
Lesser Antilles. Curagao (Miller, 1776); Aruba (H. V. Minton). Venezuela.
Br. Guiana: Kartabo, Bartica Distr.; Dadanawa (Terry, Holden); Upper Ireng Riv.,
Pakaraima Mts. (Myers). Ecuador: Galdpagos Is. Charles Isl. (H. H. Cleaves);

42 HERBERT W. LEVI
Floreana. Paraquay. (Badcock, 1932); Apa. Chile. (Nicolet, 1849); Santiago;

Talcahuano. Argentina. Santiago del Estero: Colonia Dora (B. A. Houssay).
Chaco: Gral. Pinedo (G. Cranwell). Patagonia: Santa Cruz; Rio Coyley (B.
Brown). Porterillos.
LATRODECTUS HYSTRIX SIMON

Figures 13-14, 29-30, 32; Map 7
Latrodectus hystrix Simon, 1890, Ann. Soc. Ent. France, (6) 10: 99.
Female type from Aden, in the Museum National d'Histoire Naturelle,
Paris. Roewer, 1942, Katalog der Araneae, 1: 426. Bonnet, 1957,
Bibliographia Araneorum, 2: 2372.
DIAGNOSIS.The abdomen of females examined (Figs. 29, 30) is sub-
triangular, having many short spines, but no long setae. The abdomen
seems to be dark with many light spots. The ducts of the internal genitalia
are very short (Figs. 13, 14).
The male is not known but it is expected to have a short embolus.
HABITAT. Reported from base of plants and rocks close to the ground
(Simon, 1890).
RECORDS. Aden and Yemen.
Latrodectus dahli Sp. Nov.

Figures 11-12; Map 7
TYPE. Female holotype from Bushire, Iran in the British Museum,
Natural History.
DIAGNOSIS. This species is similar to L. mactans, differing in shorter
connecting ducts (Figs. 11, 12). The ducts are wider than those of L.
hystrix.
DESCRIPTION. Sclerotized portions brown, abdomen black without
any marks. Eyes subequal in size. Anterior eyes separated by one and
one-quarter diameters. Posterior medians separated by one and one-
quarter diameters, one and three-quarters diameter from laterals. Laterals
separated by one and one-half diameters. Carapace with a median
depression. Dorsum of abdomen covered with strong long spines, and
smaller finer spines (as in L. mactans from Nepal or Peru). Epigynum
typical, opening three and one-half times as wide as long, on a raised area
with a slight transverse groove on each side of opening. Measurements:
Total length of female holotype 11.2 mm. Carapace 4.4 mm. long, 4.3
mm. wide. First femur 5.9 mm. Second femur 6.5 mm.; patella and
tibia 4.8 mm.; metatarsus 4.5 mm.; tarsus 1.4 mm. Third patella and
tibia 4.3 mm.; fourth, 6.3 mm.
This species is named for Friedrich Dahl.
RECORDS: Iran: Bushire, two female paratypes. Sokotra: Tamarida.

LITERATURE CITED
BADCOCK,H. D. 1932. Arachnida from the Paraguayan Chaco. Jour. Linn. Soc.
London, 38: 1-48.
BIANCHI,F. A. 1944. Notes on the abundance of the spiders Latrodectus mactans,
L. geometricus. Proc. Hawaiian Ent. Soc., 12: 245-247.
BONNET, P. 1945,1957. Bibliographia Araneorum, Toulouse, 1(1945);2(pt.3) (1957).
BUCKLEY, E. E. and PORGES, N. edit. 1956. Venoms, Washington, D. C.
CAMBRIDGE, F. P. 1902. On the spiders of the genus Latrodectus Walckenaer. Proc.
Zool. Soc. London, 1: 247-261.
1902a. On the genus Latrodectus. Ann. Mag. Nat. Hist., ser. 7, 10: 38-40.
CHAMBERLIN, R. V. and IVIE, W. 1935. The black widow spider and its varieties
in the United States. Bull. Univ. Utah, Biol. Ser., 3(1): 1-29.
1940. Agelenid spiders of the genus Cicurina. Bull. Univ. Utah, Biol. Ser., 5(9):
1-108.
1941. North American Agelenidae of the Genera Agelenopsis, Calilena, Ritalena,
and Tortolena. Ann. Ent. Soc. Amer., 34: 585-628.
CROME, W. 1956. Taranteln, Skorpione und Schwarze Witwen, Wittenberg
Lutherstadt.
DAHL, F. 1902. Uber abgebrochene Copulations-organe mannlicher Spinnen im
Korper der Weibchen. Sitzungsber. naturf. Freunde Berlin, p. 36-45.
1902a. Kann ein Systematiker auch zu sorgfaltig arbeiten? Zool. Anz., 25:
705-708.
GERSCHMAN,B. S. and SCHIAPELLI, D. E. 1943. Revisi6n del g6nero Latrodectus
in Sampayo, Latrodectus mactans y Latrodectismo, p. 1-23.
GRANT, V. 1957. The plant species in theory and practice. In: E. Mayr, edit.,
The species problem, pp. 39-80.
HOLM, A. 1956. Notes on Arctic spiders of the genera Erigone and Hilaira. Ark.
Zool., (2)9: 453-467.
KEEGAN, H. L. 1955. Spiders of genus Latrodectus. Amer. Midland Nat., 54:
142-152.
LEVI, H. W. 1955. The spider genera Coressa and Achaearanea. Amer. Mus.
Novitates, 1718: 1-33.
1957. The spider genera Enoplognatha, Theridion and Paidisca. Bull. Amer.
Mus. Nat. Hist., 112: 1-123.
1957a. The spider genera Chrysso and Tidarren. Jour. New York Ent. Soc.,
63: 59-81.
1957b. The spider genera Crustulina and Steatoda. Bull. Mus. Comp. Zool.,
117: 367-424.
MAYR, E. 1957. Difficulties and importance of the biological species concept.
In: E. Mayr, edit. The species problem. Washington, D. C.
ROEWER,E. F. 1942. Katalog der Araneae, Vol. 1, Bremen.
SHULOV, A. 1940. One the biology of two Latrodectus spiders in Palestine. Proc.
Linn. Soc. London, 152: 309-328.
SIMON,E. 1892. Histoire naturelles des Araign6es, Paris, Vol. 1.
SMITHERS, R. H. 1944. Contributions to our knowledge of the genus Latrodectus.
Ann. South African Mus., 36: 263-312.


The Spider Genus Latrodectus (Araneae, Theriididae) - Herbert W. Levi, 1959.

Uploaded by

Copyright:

Available Formats

The Spider Genus Latrodectus (Araneae, Theriididae) - Herbert W. Levi, 1959.

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Spider Genus Latrodectus (Araneae, Theriididae) - Herbert W. Levi, 1959.

Uploaded by

Copyright:

Available Formats

The Spider Genus Latrodectus (Araneae, Theridiidae)

Author(s): Herbert W. Levi

This content downloaded from 129.72.2.27 on Mon, 30 Sep 2013 17:48:37 PM