175 MAJOR F. WALL, LM.S., C.M.Z.S8. 
notes I use the following formula: In fig. 28, Lab. 8; 12 (3 4 2) @, the bracketted 
figures implying contact with the eye. To take another example of the tendency 
to variation in these shields see figure 66. Here these shields are 9, the an- 
terior 7 entire, the fourth, fifth and sixth touching the eye. In figure 67B 
representing the same species, the third, fourth, fifth and sixth shields are 
divided, the upper portions of the fourth and fifth being confluent. I would use the 
formula ro, 122 (‘4 © £), and in so doing imply that three labials touch the eye, though 
in reality only two do so. It seems to me the only reasonable way of recording it. 
Unfortunately many herpetologists have taken a different view, and on the strength 
of their view created new species on grounds to my mind quite unjustifiable. To take 
a good case as illustration see figures 39 and 40. Mr. Boulenger presumably on the 
assumption that the posterior maxillary teeth in nigrocincta are not grooved (though 
this is a mistake) compared the specimen he subsequently described as hendersoni only 
with species he had tabulated as Distiva, not heeding the many extremely close 
affinities this specimen bears to nigrocincta. In describing the specimen he calls the 
upper part of the divided second supralabial a loreal, the upper part of the third a 
preeocular, and the upper parts of the fourth and fifth suboculars. He says that no 
labial touches the eye onthe left side, and only the fourth on the right side. Now 
it appears to me obvious that the supralabials should be considered as follows on the 
left side : 8, 12 (2, 4, 2), 8. On the right side in this specimen they are 8, 13 (2, 4 2) &. 
On both sides three shields touch the eye. A comparison of these figures side by side 
with those of P. cyanocincta and A. stokest shows how complete is the analogy. In 
the majority of species the third and fourth supralabials touch the eye with great 
constancy, though they may be divided or not; in many species, however, examples 
are to be found in which the fifth also finds contact. The result is that with the one 
or two exceptions first noted these shields do not assist classification in any way. 
TTEMPORALS.—These shields have been conceded, I consider, undue prominence in 
classification ; for although it is true that a single large anterior shield is to be seen 
with great constancy in many of the species including many of the genus Distira, such 
as gracilis, cantoris, fasciata, obscura, etc., it is equally true that in many of the species 
of Distira especially, these shields present in many individuals departures from the 
normal. As in the case of pr@ocu/ars and postoculars, the number of these shields 
depends, to a large extent, upon the tendency of the supralabials to subdivision, for 
many herpetologists regard as lower temporals what appear to me to be the upper 
parts of divided supralabials. I find, however, that even when these shields are viewed, 
as I regard them, they vary considerably in the individuals of many species, and 
their value has, I think, been overrated. There are some instances of an abnormal 
condition in these shields prompting the creation of a new species. 
INFRALABIALS.—I regard as infralabials only those enlarged shields which are in 
contact with the sublinguals. They are distinctive and of generic value in one 
instance, viz., Emydocephalus, where the second of the series is a very long shield 
bordering most of the lower lip (see fig. 4B) ; specifically their value is but limited. In 
