The differences between this species and Pseudoeurycea gigantea are minor. Taylor (1939a) distinguished gigantea from belli by the larger size, fewer intercostal spaces between adpressed limbs, more vomerine teeth, and absence of occipital spots in gigantea. Taylor and Smith (1945) stated that in life the spots in gigantea are orange instead of red as in belli. Five specimens of Pseudoeurycea belli from Michoacán, including one juvenile, lack occipital spots. In the 34 living individuals that I have seen from Michoacán the spots varied from deep red to orange. Therefore, of the characters listed by Taylor (op. cit.) to diagnose Pseudoeurycea gigantea, only the over-all larger size and smaller number of intercostal spaces between adpressed limbs (= relatively longer limbs) are useful in separating Pseudoeurycea belli and gigantea.

Pseudoeurycea robertsi (Taylor)

Previously this species has been recorded only from the type locality. In July, 1956, individuals referable to this species were found at two sites in pine-fir forest immediately to the east of Macho de Agua and in pine-oak-fir forest at Atzimba. On August 20, 1958, a series was collected in pine-fir forest at Puerto Lengua de Vaca. These localities are between 2900 and 3000 meters in the Cordillera Volcánica in eastern Michoacán.

In life the coloration of these salamanders was highly variable. The belly and undersurfaces of the tail and hind limbs were pale gray, with or without silvery white flecks; the chin was a cream-color and flecked with silvery white in some specimens. The middorsal area was brown, orange-brown, or dull grayish yellow. The flanks and lateral surfaces of the tail were black with yellowish flecks or streaks on the flanks and yellowish or orange-brown flecks on the tail. The iris was golden brown. Measurements of eight males and two females are, respectively: snout-vent length, 42.5-56.0 (49.5), 54.0-60.0 (57.0); tail length, 42.0-56.0 (48.1), 52.0-55.0 (53.5). The smallest juvenile has a snout-vent length of 28.0 mm. and a tail length of 23.0 mm. Of the 26 available specimens, six have 12 costal grooves, and the others have 11.

In comparison with 36 topotypes, the specimens from Michoacán have a less striking dorsal color pattern; none has a well-defined dorsal reddish brown area or bold reddish mottling on the tail. Furthermore, the specimens from Michoacán have paler venters than do topotypic specimens.

Salientia

Rhinophrynus dorsalis Duméril and Bibron

These specimens (BMNH 1914.1.28.181-90) were collected by Gadow in 1908 and reported by him (1930:72): "Whilst this very sluggish termite-eating toad is common enough in the sweltering hot country of the state of Vera Cruz, up to an elevation of 1500 feet, it was unknown on the west side of the Isthmus until I found it in great numbers near the mouth of the Balsas River, in and near fresh-water pools, where it attracted attention by its loud peculiar voice during the pairing season in the month of July." Subsequently, Peters (1954:3) verified the identification of these specimens. Although torrential rains fell during the week in July, 1955, that I spent at Playa Azul near the mouth of the Río Balsas, the distinctive voice of Rhinophrynus was not heard. Elsewhere on the Pacific coast of México adult Rhinophrynus have been reported only from Tehuantepec and a few localities on the coastal lowlands of Chiapas. Taylor (1942b:37) found on the coast of Guerrero a tadpole that was referred to the genus Rhinophrynus by Orton (1943). In the summer of 1960 adults of Rhinophrynus were collected near Acapulco, Guerrero (Fouquette, in litt.). These recent collections verify the existence of the species along the Pacific lowlands of México at least as far north as Michoacán.

Scaphiopus hammondi multiplicatus Cope

This small toad has been found at elevations between 1500 and 2500 meters on the Mexican Plateau and associated mountain ranges; it occurs in mesquite-grassland and in pine forests. Calling males and females laden with eggs have been collected in the rainy season in the months of July and August. The call is a medium-pitched snore. In living individuals the dorsal ground color varies from pale brown to gray with dark brown or olive-brown markings. In many individuals the tips of the small dorsal pustules are red.

Bufo coccifer Cope

In life the dorsal color pattern consists of a yellowish tan ground color with dark brown spots; the middorsal stripe is deep yellow or cream color. The venter is a dusty cream color, and the iris is pale gold. Males have dark brown horny nuptial tuberosities on the thumb. The following measurements are of 21 males and four females, respectively: snout-vent length, 43.5-51.7 (48.1), 55.6-62.6 (59.1); tibia length, 16.6-18.8 (17.6), 18.8-20.3 (19.3); head width, 16.7-19.7 (18.4), 20.6-22.2 (21.4); head length, 13.8-16.6 (14.8), 16.5-18.2 (17.3).

The specimens from the Tepalcatepec Valley differ slightly from specimens from southeastern México and Central America. Those from Michoacán have low and narrow cranial crests; in about one-half of the specimens the occipital crest exists only as a row of tubercles, and in some the postorbital and suborbital crests are barely discernible. Specimens from the southern part of the range, Costa Rica and Nicaragua, have much higher and thicker cranial crests; in these the occipital crest is well defined and extends posteriorly to a point back of the anterior edge of the parotid gland; the postorbital and suborbital crests are well marked. Of 48 specimens from Esquipulas, Guatemala, all have high crests, but these are not so well developed as in ten specimens from Matagalpa, Nicaragua, and three from various localities in Costa Rica. Six specimens from Tehuantepec, Oaxaca, have cranial crests that are lower than those in specimens from Guatemala. In three of the specimens from Tehuantepec the occipital crests are reduced to a series of tubercles. Of six specimens from Agua del Obispo, Guerrero, four have poorly developed occipital crests. These observations suggest the presence of a cline in the development of the cranial crests; specimens have higher crests in the southern part of the range than in the northern part.

In México Bufo coccifer has been collected only in semi-xeric habitats, but to the south, from Guatemala to Costa Rica, it has been found in more upland and humid habitats. Southern specimens are darker than those from the north, a possible correlation with the differences in habitat.

These toads probably range throughout the Tepalcatepec Valley, but they are unknown from the coast of Michoacán. Breeding choruses were found after heavy rains on June 24, 1955, and on August 2, 1956. The first was in a muddy ditch; the second was in a flooded grassy field. The call is a high-pitched, but not loud, "whirrr." Males were calling from the edge of the water or from clumps of grass in the water. Clasping pairs were in the water; amplexus is axillary.

Bufo compactilis compactilis Wiegmann

The southwestern terminus of the range of this species is on the Mexican Plateau in Michoacán. All specimens from the state have spotted venters. In living toads the dorsal ground color was gray or grayish tan with olive green spots. The vocal sac was brownish gray; the iris was a bright golden color.

On June 11, 1958, many individuals were calling from shallow water in a flooded field at Emiliano Zapata. The call is a slow trill, in which the individual notes are discernible.

Bufo marinus (Linnaeus)

This large toad is characteristically found in areas supporting tropical scrub forest to elevations of about 1000 meters. The species is much more abundant than the numbers listed above suggest. In the dry season individuals have been observed in patios, along streams, and by irrigation ditches. In the rainy season the loud, rattling call of the males is heard at night throughout the Tepalcatepec Valley and the coastal lowlands.

Taylor and Smith (1945:552) revived Wiegmann's Bufo horribilis for the large toads of México that are here referred to B. marinus. Their action was based upon the supposition that the "species marinus" is composite. Although probably true, this supposition has yet to be proved. Until the large, and apparently related, species of Bufo inhabiting tropical America have been studied systematically as a unit, the recognition of segments of the population as either species or subspecies is meaningless. Taylor and Smith (op. cit.:553) based the description of a new species, Bufo angustipes, on one rather emaciated, formalin-hardened female from La Esperanza, Chiapas. The type (USNM 116513), when compared with numerous specimens of Bufo marinus from throughout the range of the species in México and northern Central America, displays no combination of characters to set it off from the others. Therefore, I suggest that Bufo horribilis Wiegmann and Bufo angustipes Taylor and Smith be placed in the synonymy of Bufo marinus (Linnaeus) until future systematic study of the genus and this species in particular establishes the existence of recognizable taxa.

Bufo marmoreus Wiegmann

In Michoacán this species is confined to elevations of less than 1000 meters on the coast and foothills of the Sierra de Coalcomán. In this region in the months of June and July, breeding congregations have been found in temporary pools and along streams.

Smith and Taylor (1948:39), in their key to the Mexican species of Bufo, placed emphasis on the nature of the supraorbital and postorbital crests (whether they form a curve or a sharp angle) in distinguishing Bufo marmoreus from Bufo perplexus. In the original description of perplexus, Taylor (1943a:347) characterized the species as follows: supraorbital and postorbital crests forming a sharp angle, instead of a curve as in marmoreus; supratympanic crest smaller than in marmoreus; diagonal lateral stripe lacking in females; concentration of dorsal tubercles as found in marmoreus lacking in males. The discovery of specimens in which the crests form a curve and others in which the crests form an angle in both the Tepalcatepec Valley and in the coastal lowlands prompted an investigation of these characters and others throughout the ranges of the species. An examination of 410 specimens has resulted in the following conclusions.

Table 1.—Variation in the Shape of the Supraorbital and Postorbital Cranial Crests in Bufo marmoreus and B. perplexus.

PLATE 1

PLATE 2

PLATE 3

PLATE 4

PLATE 5

PLATE 6

Therefore the nature of the cranial crests is of little value in separating two populations, but the color pattern of the females and the nature of the dorsal tubercles of the males do show distinct differences. Furthermore, certain differences in size and proportion are evident; Bufo marmoreus is a slightly larger toad and has a relatively longer tibia and longer head than perplexus (Table 2).

Table 2.—Comparison of Certain Measurements and Proportions in Bufo marmoreus and B. perplexus. (Means Are Given in Parentheses Below the Ranges.)

Taylor (1943a:347) described Bufo perplexus from Mexcala on the Río Balsas in Guerrero. Among the many paratypes are specimens from Tonolá, Chiapas, and Tehuantepec, Oaxaca. These apparently were referred to perplexus solely on the nature of the cranial crests. All of the specimens examined during the course of the present study from the lowlands of Veracruz and from the Pacific lowlands from Sinaloa southward to Chiapas are referable to Bufo marmoreus; those from the Balsas-Tepalcatepec Basin are referable to Bufo perplexus, as defined above. Ten specimens from Chilpancingo, Guerrero (UMMZ 115352), do not readily fit either species. Perhaps there is gene exchange between the inland and coastal populations through the relatively low pass at Chilpancingo, at the mouth of the Río Balsas, and near the convergent headwaters of the Río Coahuayana and Río Tepalcatepec in southern Jalisco. If this can be demonstrated, then Bufo perplexus would have to be considered as a subspecies of Bufo marmoreus, instead of an allopatric species.

Bufo perplexus Taylor

Bufo occidentalis Camerano

This toad is an inhabitant of pine and oak forests between 900 and 2400 meters. Near Charapendo on the slopes of the Sierra de los Tarascos and at Coalcomán it apparently reaches its lowest altitudinal limits. At both of these localities the pine-oak forest is replaced by arid tropical scrub forest on the lower slopes.

Twenty-four tadpoles were collected on May 3 in a quiet section of a fast stream near Barranca Seca. The tadpoles have a robust body, broadest about two-thirds the distance from the snout to the posterior edge of the body, half again as broad as deep. Eyes dorsolateral; nostrils dorsal, somewhat directed forward, and about three-fifths the distance from the tip of the snout to the eye; spiracle sinistral and lateral, located at about midbody; anus median; tail long and slender; tail-musculature extends nearly to tip of tail; depth of tail-musculature at mid-length about one-third total depth of tail; dorsal tail-fin not extending onto body (Fig. 5); average body length of ten tadpoles having small hind limb buds, 14.4 mm.; average tail length, 22.0 mm.

Mouth ventral, nearly terminal, about one-third as wide as widest part of body; anterior lip has no papillae; lower lip bordered by two rows of papillae and lateral lips by one row of papillae; beaks moderately well developed, the upper forming a broad arch and finely denticulate; tooth rows 2/3, the upper rows extending to the edge of the lips, subequal in length, and slightly longer than lower rows, which also are subequal in length; inner upper tooth row broken medially; inner lower tooth row sometimes broken (Fig. 6).

The body is black dorsally and laterally, and bluish gray ventrally; the tail musculature is brown and stippled with darker brown. The fins are transparent and stippled with brown, the stippling being most pronounced on the posterior two-thirds of the upper tail-fin.

Forty recently metamorphosed individuals average 18.9 mm. in snout-vent length.

The relationships of this toad seem to be with Bufo bocourti Brocchi, an inhabitant of pine and oak forests in the uplands of Chiapas and Guatemala. In Bufo occidentalis the tympanum usually is indistinct and sometimes completely covered, and it is absent in bocourti. Bufo occidentalis has a broader interorbital area and relatively shorter and more rounded parotid glands than bocourti. The tadpoles of the two species are nearly identical (see Stuart, 1943:12).

Leptodactylus labialis (Cope)

In the Tepalcatepec Valley this frog reaches the northernmost known limit of its range in western México. Although the species is abundant in the valley, it apparently is absent from the coastal lowlands. In the Tepalcatepec Valley Leptodactylus melanonotus seems to be more abundant than labialis. In the rainy season both species have been heard calling from the same ponds and flooded fields.

There are only slight differences in size between the sexes; measurements of 20 males and eight females are, respectively: snout-vent length, 32.3-39.5 (35.1), 34.1-39.2 (37.2); tibia length, 14.3-17.0 (15.4), 14.9-16.8 (15.8); head width, 11.0-13.6 (12.0), 12.2-13.2 (12.6); head length, 12.8-15.1 (13.3), 12.8-14.6 (13.7).

Leptodactylus melanonotus (Hallowell)

This species is widespread in the lowlands of the state; it has been collected up to elevations of 1050 meters in the Tepalcatepec Valley. In the dry season individuals were discovered beneath rocks along streams and in damp arroyos; in the rainy season they were found wherever there was water. Males were heard calling from flooded fields, ditches, rocky streams, and small puddles. The call is a series of individual notes: "woink, woink, woink."

Adult males are noticeably smaller than females; measurements for 20 males and ten females from Apatzingán are, respectively: snout-vent length, 29.6-34.6 (32.3), 36.3-44.1 (40.8); tibia length, 12.6-15.1 (14.0), 16.5-19.0 (17.8); head width, 10.8-11.9 (11.3), 12.6-14.8 (13.7); head length, 11.2-13.2 (11.9), 13.1-14.8 (14.0). Brownish yellow ventral glands are present in some juveniles and in some adults collected in the dry season as well as in the rainy season.

Leptodactylus occidentalis Taylor

On the night of June 11, 1958, this species was calling from a hyacinth-choked ditch. Although numerous individuals were heard, only one specimen was obtained. The frogs were calling from the tangled mat of hyacinths along with Hyla eximia, Hypopachus oxyrrhinus ovis, and Rana pipiens.

Taylor (1936a:352) characterized this species as follows: "The narrow head, small maximum size (38 mm. for females, 33 mm. for males), the character of the postaxillary and postfemoral glands, the narrower groups of vomerine teeth, clearly distinguish this western Mexican form from the more robust, larger melanonotus to the south. The call is likewise fainter and different in quality." Concerning the glands, Taylor (loc. cit.) remarked: "There is a possibility that the horny excrescence covering the glands may appear only during the breeding season. This character is quite as strongly marked in females as in males." Bogert and Oliver (1945:324) concluded that the population of Leptodactylus in northwestern México could not be distinguished from melanonotus in other parts of the country and thus synonymized Leptodactylus occidentalis with melanonotus. Bogert and Oliver (op. cit.: 324) stated that the extent as well as the presence or absence of ventral glands was highly variable in all samples examined by them.

Upon seeing numerous living individuals of Leptodactylus melanonotus from many parts of its range in México and individuals of the population of Leptodactylus in northwestern México (Nayarit and Sinaloa), I was immediately impressed not so much by the differences in the development of the ventral glands, but by the color of the glands. The differences in color are apparent in freshly preserved specimens. With the exception of Leptodactylus from northwestern México, specimens of melanonotus from throughout México and northern Central America have yellow or yellowish brown glands. Specimens from northwestern México have black or brownish black glands that are conspicuously darker than those found in melanonotus. Examination of 653 preserved specimens of Leptodactylus melanonotus from México and Guatemala has failed to reveal specimens with black ventral glands, like those found in specimens from northwestern México, to which the name Leptodactylus occidentalis has been applied. Furthermore, in melanonotus the glands are less distinct and more extensive than in occidentalis; in the latter species glands are absent from the throat and midventral area, where they often are present in melanonotus (Fig. 7).

In some individuals of both species collected in the dry season and in some collected in the rainy (breeding) season the glands are absent; the development of these glands, therefore, does not seem to be correlated with breeding. Likewise, the glands are present or absent in either sex, and often as not they are present in juveniles. Presence of the glands, therefore, cannot be correlated either with sexual or ontogenetic development. Since the glands are found in individuals from all parts of the range, it is unlikely that there is a correlation between the development of the glands and the environment.

Aside from the differences in the ventral glands, the call is different in the two populations. The call of Leptodactylus occidentalis is a rather harsh "wack, wack, wack" as contrasted with the more nasal "woink, woink, woink" of melanonotus. Sound spectrographs are needed to analyze the differences in calls. None of the specimens of occidentalis examined approaches in size the largest individuals of melanonotus; possibly the size of the frogs is another valid character for separating the species. On the basis of the above data it is evident that the frogs in northwestern México show certain characters that distinguish them from Leptodactylus melanonotus, as it is known throughout the rest of México. It is not known for certain that melanonotus and occidentalis are sympatric. Several series of old, poorly preserved specimens from Nayarit and Sinaloa cannot be placed in either species, for none has visible ventral glands. Leptodactylus melanonotus is known from Acaponeta, Nayarit (AMNH 43913-25), and the following localities in Jalisco: Barro de Navidad (UMMZ 118098), La Concepción (UMMZ 113081), La Resolana (UMMZ 102104), and Tenachitlán (UMMZ 113045-6). Records for Leptodactylus occidentalis are: Álamos, Sonora (AMNH 51356-65); Culiacán (AMNH 49511-9), Chele (UMMZ 110914), and Rosario (UMMZ 113062) in Sinaloa; Ixtlán del Río (UMMZ 102108), San Blas (UMMZ 112814, 112994, 110892, 115543), and Tepic (UMMZ 115544) in Nayarit; Ameca (UMMZ 102106-7) and La Cofradía on the south shore of Lago de Chapala (UMMZ 102105) in Jalisco; and Tangamandapio, Michoacán (UMMZ 119145). From these scattered records it appears that Leptodactylus occidentalis in the southern part of its range stays in the uplands, whereas melanonotus is confined to the lowlands.

Microbatrachylus hobartsmithi (Taylor)

Of six specimens from Cascada Tzararacua, five are colored like typical M. hobartsmithi, having the anterior and posterior surfaces of the thighs and the upper arms pale pink in life and a grayish brown dorsum in preservative. The other specimen (UMMZ 94231) has in preservative a dark brown dorsolateral line on each side enclosing a pale tan area that extends from the snout to the vent. One specimen from 29 kilometers east of Morelia (UIMNH 40338) and 13 specimens from San José de la Cumbre (UMMZ 102111) do not have the prominent tarsal tubercles characteristic of M. hobartsmithi. Also, in these fourteen specimens the palmar tubercles are larger, and the dark anal patch more distinct, than in typical M. hobartsmithi. Possibly these specimens, which are from the high mountains in the eastern part of Michoacán, represent another species of Microbatrachylus. However, Taylor (1940d:501) reported a series of M. hobartsmithi from the mountains 10 miles west of Villa Victoria in the western part of the state of México.

The largest specimen from Michoacán is a gravid female (UIMNH 16104) having a snout-vent length of 23.5 mm.

Microbatrachylus hobartsmithi has been found in rocky ravines along streams in the Cordillera Volcánica and the southwestern escarpment of these mountains at elevations from 1450 to 2750 meters.

Microbatrachylus pygmaeus (Taylor)

This large series (UMMZ 119247-8) was collected on June 22 and 23, 1958, before the onset of the heavy summer rains. The frogs were found in a shaded ravine at the north edge of Arteaga; they were obtained during the day, at which time they were actively moving about in the leaf litter along a small stream.

These frogs are all referred to M. pygmaeus, because this is the earliest name available for frogs showing the variation in characteristics displayed by this large series. The characters used by Taylor (1936a, 1940d, 1941a, and 1942b) and Smith and Taylor (1948) to distinguish the various species of Microbatrachylus include color pattern, relative length of the hind limb, presence and position of dorsal dermal folds or pustules, relative size of inner and outer metatarsal tubercles, and the number of palmar tubercles. All specimens from Arteaga have two palmar tubercles; the inner and outer metatarsal tubercles are subequal in size. Furthermore, aside from sexual difference, there is little variation in the relative length of the hind limbs (Table 3). However, many color patterns do exist in the series; each of these color patterns is described below.

Table 3.—Snout-vent Length Expressed as a Percentage of Tibia Length in Animals of Six Color Patterns of Microbatrachylus pygmaeus. (Letters Refer to the Variants Having the Color Pattern Discussed Immediately Below)

A.—225 specimens: Dorsum mottled brown and cream, usually with a dark spot between the eyes and one or two dark V-shaped marks with the apex anteriorly on the back; 55 of these have a narrow cream-colored line from the tip of the snout to the vent and thence onto the posterior surfaces of the thighs. All are pustulate above; in most specimens the pustules form no pattern, but in some they tend to form a V in the scapular region.

B.—41 specimens: Dorsum pale tan or cream-color with brown mottling on flanks; a brown interorbital bar and a brown chevron in scapular region. Dorsum irregularly pustulate; in some specimens the pustules tend to form a V in the scapular region.

C.—12 specimens: Dorsum colored like "A", but having a broad yellow stripe narrowly bordered by black from the tip of the snout to the vent; in some specimens there is a narrow yellow stripe on the posterior surfaces of the thighs. The dorsum is irregularly pustulate.

D.—31 specimens: Dorsum variably streaked with cream-color or pale tan and brown; usually a broad cream-colored stripe from eyelid to groin bordered laterally by a somewhat narrower brown stripe; middorsal area cream-color and separated from dorsolateral cream-colored stripe by a brown stripe, or middorsal area brown with a cream-colored or yellow, narrow stripe from tip of snout to vent; a dark stripe from tympanum to flank; dorsal surfaces of heels creamy white to pale orange; anal patch brown. A dermal ridge from posterior edge of eyelid to rump; another ridge extends posteromedially from the eyelid; scattered pustules on the dorsum in some specimens.

E.—17 specimens: A narrow dark stripe from snout, through nostril and eye, over tympanum, to vent, enclosing a unicolor dorsum (reddish tan to yellowish tan in life); heels pale tan or yellow above; anal patch black. A faint dermal ridge from posterior edge of eyelid to rump, or part way to rump.

F.—2 specimens: Mottled brown and cream-color above; upper lips and upper arms white. A dermal fold from posterior edge of eyelid to rump; scattered pustules on dorsum.

Some of these color variants are assignable to names proposed by Taylor: "A" and "B" undoubtedly are M. pygmaeus (Taylor, 1936a); "C" probably is M. pygmaeus; "D" is referable to M. minimus (Taylor, 1940d) in most characteristics, although the coloration is more nearly like that of M. lineatissimus (Taylor, 1941a), a larger species characterized by a relatively long hind limb; "E" apparently is M. imitator (Taylor, 1942b); "F" is M. albolabris (Taylor, 1940d). Examination of series of these frogs from other parts of México shows a similar composition of color variants. Of 78 specimens from the Río Sarabia and the village of Sarabia in Oaxaca (UMMZ 115428-37), 57 are "A," six are "D," three are "E," and 12 are "F"; of 22 specimens from Teapa, Tabasco (UMMZ 113829), 11 are "A," five are "D," two are "E," and four are "F"; of 33 specimens from Potrero Viejo, Veracruz (USNM 115447-58, 115461-71, 116840-2, 116864-70), ten are "A," 13 are "E," and ten are "F"; of 31 specimens from La Esperanza, Chiapas (USNM 115477-9, 116827-39, 116849-63), 28 are "A" and four are "F."

It is highly doubtful if these color variants are actually distinct species. Goin (1950 and 1954) in his studies of inheritance of color pattern in West Indian species of the genus Eleutherodactylus has shown that similar color pattern variants come from the same clutch of eggs; furthermore, Goin has worked out the genetic ratios of certain of these variants. Heathwole (in litt.) obtained "normal" specimens and individuals having a broad middorsal stripe ("C" in figure 9) from a clutch of eggs of Eleutherodactylus gollmeri. The presence of a broad middorsal yellow stripe is common in Eleutherodactylus rugulosus.

Perhaps the most interesting aspect of variability in color pattern in Mexican eleutherodactylids is the parallelism between members of the Eleutherodactylus rhodopis-group and some members of Microbatrachylus. In the former group there are white-lipped individuals (Eleutherodactylus beatae Boulenger), individuals having a unicolor reddish or yellowish dorsum (E. dorsoconcolor Taylor), and individuals having a dorsal pattern of irregular longitudinal brown and cream-colored streaks (E. venustus Günther). In the humid forests of southern Veracruz, northern Oaxaca, and Chiapas members of both groups occur sympatrically. A proper understanding of the evolutionary significance of these variants in the two groups, as well as proper allocation of the presently recognized species, must await experimental evidence based on studies of the inheritance of color pattern. Nevertheless, at present it is apparent that certain characters, especially the nature of the dermal folds and pustules, and the color pattern, are of little taxonomic value in distinguishing "species" of Microbatrachylus. The data derived from a study of the large series from Arteaga, together with that from the other series examined, suggests that Microbatrachylus albolabris, imitator, minimus, and pygmaeus are morphotypes of one species. Of these names, pygmaeus is the oldest. Consequently Microbatrachylus pygmaeus has been used here for the series from Arteaga.

Although Microbatrachylus hobartsmithi, a species distinguished from all of the above by the presence of tubercles on the outer edge of the tarsus, is known from Michoacán northward into Nayarit, Microbatrachylus pygmaeus previously has not been known north of Guerrero, where it occurs in habitats similar to that in which it was collected at Arteaga.

Eleutherodactylus augusti cactorum Taylor