58 THE EEPTILES OF EGYPT. 
from Socotra; fifth, P. crucifer, Y&\., from Berbera (Las Gori), Abyssinia, Western 
Somaliland, and S. Arabia ; sixth, P. phillipsii, Blgr., also from Berbera; and, lastly, 
P. carteri, Gray, from the island of Masira off the east coast of Arabia. Another 
species was described by Steindachner as Spatalura collaris, but the locality whence it 
was obtained is unknown. Mr. Boulenger provisionally regards it as a variety of 
P. carteri, Gray. 
The species of Pristurus are divided by Mr. Boulenger into two groups, depending 
on whether the rostral shield enters into or does not enter into the formation of the 
nostril. In the first four the rostral borders the nostril, while in the remaining three 
species it does not. 
P. percristatus and P. flavipunctatus are closely allied, but the males of the former 
are easily distinguished from those of the latter by their stronger dorsal crest extending 
to the nape, whereas in P. flavipunctatus it extends on to the anterior half of the body 
only. These stronger-crested geckos have a slightly longer hind limb than P. flavi- 
punctatus, as it may extend to the shoulder or between the shoulder and the ear, 
while in P.flavipunctatus it reaches the axilla or slightly beyond the shoulder. The 
variability of the crest in Pristurus, the extent to which the length of the limbs varies 
in P.flavipunctatus, and the difficulty that will be experienced in distinguishing the 
females of P. percristatus from those of P. flavipunctatus are considerations which 
suggest the possibility of the former being only a variety of the latter. Characters 
derived from the degree of development of the crest seem to be misleading, for 
when Mr. Boulenger defined P. crucifer and P. phillipsii his materials led him to 
conclude that while the latter had a caudal crest, the former had none. Since then 
male specimens of P. crucifer with a very rudimentary caudal crest have been recorded 
by him from Capt. Bottego's collection from Southern Somaliland, so that P. crucifer 
has a caudal crest as well as P. phillipsii. 
I recently referred 39 specimens of a gecko of this genus from the Hadramut, 
Arabia, with considerable doubt, to the species described by Steindachner as 
P. collaris, but pointed out at the same time that they differed from P. carteri, 
Gray, the types of which were before me, in having a well-defined black collar and 
by the absence of a mesial patch of spiny scales on the middle of the belly. In the 
original description of Spatalura collaris, Stdn.=P. collaris, the lizard is said to have 
a dorsal crest, a structure entirely absent in the Hadramut geckos. Some of the 
latter, however, were compared by Dr. Werner with the types preserved in the Vienna 
Museum, with which he found them to be perfectly identical, so that I was entitled 
to conclude that the types had no dorsal crest, and that an error had crept into 
Steindachner's description. If, however, on further examination the types of P. collaris 
should reveal the presence of a rudimentary dorsal crest, its absence in the Hadramut 
specimens would not necessarily imply that they were distinct from P. collaris, in view 
of the variable development of the dorsal and caudal crests in other species. 
