1896.] FISHES MADE BY DR. A. D. SMITH. 219 
4, CHROMIS SPILURUS, Giinth. Proc. Zool. Soc. 1894, p. 89. 
This species was discovered by Dr. Gregory in pools remaining 
in dried-up watercourses of North Giriama. Dr. Donaldson Smith 
found this species (30.12.94) under similar conditions near the 
Shebeli River, and (8.12.94) in water-holes near Sheikh Husein. 
All the specimens, those collected by Dr. Gregory as well as by 
Dr. D. Smith, are small, not exceeding 12 cm. in length. 
5. CLARIAS SMITHII, sp. n. 
Clarias lazera, Giinth. P. Z. 8. 1894, p. 89 (nec C. V.). 
D. 70-73. A. 58-62. V.1/9. 
Vomerine teeth (Fig. 1) granular, forming a very broad band, 
nearly twice as broad as that of the intermaxillary teeth, with an 
obtuse, rounded projection behind in the middle of its concavity. 
Transversely the intermaxillary band is wider than the vomerine. 
The mandibulary dental band is as broad as the intermaxillary. 
Upper surface of the head with not very coarse granulations; the 
length of the head is two sevenths of the total, without caudal. 
The maxillary barbel reaches beyond the root of the pectoral, the 
nasal barbel being not quite half its length. The pectoral fin 
extends to, or nearly to, the origin of the dorsal, the spine being 
two thirds of the fin. Dorsal fin separated by a short interspace 
from the caudal. 
Fig. 1. 
Teeth of Clarias smithii. 
A single specimen, 45 centim. long, is in the collection, and was 
captured in the middle course of the Shebeli. The breadth of the 
intermaxillary band of teeth is 5 millim., that of the vomerine 
8 millim.; the transverse width of the former is 51, of the latter 
47 millim. A fish captured by Dr. Gregory at Negatana, and 
enumerated by me under the name of Clarias lazera, is evidently 
of the same species. 
The African species of Clarias are extremely similar in general 
appearance and most difficult to define, chiefly on account of the 
uncertainty which attaches to almost all the taxonomic characters 
which have been used for distinguishing them. Some of the 
characters are certain to undergo considerable changes with age, 
for instance the vomerine teeth, which are not likely to be granular 
in very young specimens. The presence or absence of a posterior 
process of the vomerine band is a more reliable character, as is 
