386 PROCEEDINGS OF THE NATIONAL MUSEUM YOL, 92 
Diagnosis.—Vertebrals enlarged, about twice width of adjacent 
scales; usually three labials enter orbit; bands on body numerous, 52 
to 65, broken posteriorly or not. 
Range.—Pacific coast of Chiapas, south to Panama. 
Specimens examined.—Four. 
Remarks.—This species is distinguished from all others by having 
most of the vertebral scales about twice as wide as the adjacent scales; 
they are larger than in any other species except cenchoa. 
The name gemmistratus has long been used for the Mexican Jman- 
todes with scarcely or not enlarged vertebrals. Even Cope eventually 
used it in this sense. Unfortunately, the type is missing, so it is 
impossible to prove what form Cope may have originally described. 
Fortunately, however, the type locality is definite. It is fairly cer- 
tain that the name does not apply to cenchoa, because (1) Cope 
definitely says the vertebrals are relatively small, (2) he gives 42 
bands on body, while 40 is the maximum for leucomelas (definitely 
known from southeastern Chiapas, therefore probably the species 
of Pacific coast Salvador, which is in the same faunal zone), and (3) 
he says the bands “are peculiar in being connected by a median dorsal 
vitta.” The latter is a characteristic of many specimens of other 
species of Jmantodes, but not of cenchoa. 
I conceive that gemmistratus and splendidus (see below) are mem- 
bers of a single morphologic group (having similar ventral and caudal 
scale counts, similar pattern), and accordingly that one does not occur 
with the other. If this is true, then the species with small dorsals 
(gemmistratus auct.=splendidus) cannot occur in the same territory 
with the species with larger dorsals (elegans auct.=gemmistratus). 
Since the only specimens now known from the faunal area that includes 
Izalco, type locality of gemmistratus, are two 1° which have the verte- 
bral scales enlarged (as in elegans auct., I believe the name gemmi- 
stratus must be applied to that species. Unfortunately it is an older 
name than elegans, which it replaces. 
While gemmistratus is very uniform in character of the vertebral 
scales, extensive variation in pattern occurs. Panama and Costa Rica 
specimens have the bands complete on the posterior part of the body, 
as well as anterior, while Nicaragua and Chiapas specimens have them 
broken posteriorly. The ventrals and caudals are usually low (as in 
Mexican splendidus), and vary from 220 to 237 (ventrals) and 114 to 
146 (caudals). The exception is a small Nicaragua specimen with 251 
ventrals and 155 caudals (U.S.N.M. No. 25248). 
19 T.a Esperanza, Chiapas (U.S.N.M. No. 110521) and Pacific Coast of Guatemala (Brit. 
Mus.). 
