Cockereil — The Scales of the African Ci/prinid Fishes. 151 



of two lots of species, one Asiatic, the other from tlie Transvaal. 

 The Asiatic inemhers are L. sladoul (Mandalay, F. Day), L. 

 ricnorhiinchiis (Janiu, ITiinalayas, SchJnijiiitireit), L. biculor (Cal- 

 cutta, F. Dai/), L. s///(/t'//.s/.s and L. (Jiplot^tonuis. The Transvaal 

 ones are L. ?"o.s.r, rnthli and capensls. This .(jfrouj) prohahly has no 

 proper standing, bnt consists ratlu'r of slightly divergent members 

 of the last. 



( d. ) L. aUivelis group. The often large scales sculptured about as in 

 the ?/t<7C?"Osfo?rta group, but subcircular, without laterobasal angles. 

 All African, including L. allivrlis, sencf/aloii^is, Jineatns, couhie 

 (many basal radii), darllngi and timhratus. The last is tlie tyjje 

 of Ahrostomus A. Smith. Jj. lioric forms a subgroup with reduced 

 sculpture. 



(e.) L. icalkcri (bracJiijporim) group. Type of inacrostomn, but the 

 scales becoming ()l)tusely subtriangular. Tiiis may be com|)ared 

 with tiie hrcviharhls grouj) of Ilarhus. 



( f. ) L. cJiri/sopJiekadion grouj). This species, from the jNIenam 

 River, Siam, is very distinct by the character of the apical circuli, 

 which instead of being broken up into tubercles, are continuous, 

 and meet at a sharp angle in the middle line, as in Cirrhiiia 

 jnllieiii. They are however denser than in the Cirrhina. Tj. 

 rohila from Calcutta (,,Thos. Moore) has essentially the same struc- 

 ture, but the circuli are more or less broken ui) into tubercles. 

 The shape of the scale is essentially as in the macrostoma group. 



It will be seen that the grouping of Laheo here given does not 

 accord well with that based (e. g. by Boulenger) on the fishes. 

 There is however a tendency for the groups to agree with geograph- 

 ical areas, and it is probable that they have some real significance. 

 No doubt some of the groups are artificial, including similar look- 

 ing scales which actually have difl'erent origins. 



Tiegarded as a whole, the African Cyprinid fauna i)resents some 

 curious historical problems. Africa is the home of a varied fauna 

 of Characinids, which belong to endi'mic genera. As America was 

 ])rol)ably the home of this family, and in view of the distinctness 

 of the African Characinid fauna from tliat found in the Neotropical 

 Region, it nnist be supposed that the arrival of the Characinids in 

 Africa took place very long ago, perhaps in the Mesozoic. On the 

 other hand, the Ethiopian Cyprinid fauna is obviously Asiatic, and 

 the amount of endemicity (aside from species) is so slight that we 

 must postulate a more recent period of arrival, certainly Tertiary, 

 perhaps not earlier than ^Miocene. It is curious that with the 

 country so well stocked in advance with Characinids, it was ]>ossi- 

 ble for the Cyprinids to spread even to the extreme south, and 

 produce such a large nundjer of specific forms. This may be due 

 in part to a greater adaptability (already well developed in Asia) 



