November 7, 1907] 



NA TURE 



.sented only by a single species, follow, the last being 

 Cromeria, distinguished from the Galaxiidae of the 

 Haplomi by the presence of a mesocoracoid (Swinner- 

 ton). The family of the Characinidaa (under the 

 Ostariophysi) form a very generalised group confined 

 to the fresh waters of Africa and Central and South 

 America, from which the author thinks they may have 

 migrated by a land connection in Upper Cretaceous 

 times. These supposed precursors of the Cyprinoids 

 number eighteen species in the Nile. 



The widely distributed family of the Cyprinidae com- 

 prises the largest number of species within its limts, 

 viz. fiftv, and thirty-five of these belong to the genus 

 Barbus, a large proportion, seeing that in Day's 

 " Fishes of India and Neighbouring Regions " there 

 are but seventy. Yet the genus is conspicuous by its 

 absence from the Senegal, the Gambia, and Lake 

 Chad. The author's wealth of material has enabled 

 him to clear up the synonymy of certain species, such 

 as hates coubie, yet it is doubtful if, as in Europe, 

 hvbridism mav not occur to a greater extent than is 

 at present imagined. The representatives of the genus 



Fig. 2.-Throwing-net as used on the Lower Xi!t. From '" Ihe I- 



Barbus, of which there are no fewer than twenty- 

 seven new species in the work, offer a wide field for 

 the features just mentioned, since many are very closely 

 allied, though separable, perhaps, by such points as 

 the proportions of the snout. The tropical or sub- 

 tropical Silurids are largely represented by fifteen 

 genera and forty-one species, and the habits of some, 

 such as Clarias, are full of interest, for they spend 

 the dry season in burrows in dried-up marshes, which 

 they leave at night in quest of food — both animal and 

 vegetable — using the spines of the pectorals in pro- 

 gression. The name Malapterurus has so long been 

 used that the author's change to Malopterurus jars, 

 and for similar reasons he himself does not follow 

 Starr Jordan in calling the species " Torpedo " elec- 

 tticiis. It is noteworthy that whilst the fresh-water 

 species are all genericnlly distinct from the .\merican, 

 those species which enter the sea on both shores 

 of the .Atlantic agree (e.g. .Arius). The Cichlidae, a 

 familv which presents great difficulties from the close 

 resemblances of many — e.g. those of Lake Victoria — 

 have increased, largely by the labours of Mr. Bou- 



NO, 1984, VOL. 77] 



lenger, during the last few years from twenty species 

 to 210 in -Africa, and of these eighteen belong 

 to the Nile. The author considers that the 

 forms inhabiting that great lake (Victoria) sprang 

 from a small number of original (isolated) types, and 

 were modified into a multitude of species according 

 to lines different from those followed by other colonies. 

 Only two or three of these are identical with or very 

 closely related to forms in neighbouring rivers. 



Though many interesting facts in regard to repro- 

 duction and development are incidentally noted in this 

 fine work on Nile fishes, especially in connection with 

 Mr. Budgett's investigations on Polvpterus, Pro- 

 topterus, and on the breeding of the' Mormyridse, 

 very much yet awaits the observer in this department', 

 and no more fascinating field exists, to judge frorri 

 the fragmentary knowledge available. Some, like 

 Hyperopisus bebe, attach their oval eggs to rootlets of 

 grass, and the larva? hang in thousands, like amphi- 

 bians, to the rootlets until the volk-sac is absorbed. 

 Others have floating nests 2 feet long by i foot 

 broad (Gyiniiarchus nUoHcits) for eggs 10 mm. in 

 diameter, and for larvae with long 

 gill-filarnents. .\ still larger nest 

 (4 feet in diameter) characterises 

 Hctcrotis niloticHs. the larva; of 

 which also have gill-filaments. The 

 eggs of Cyprinodon fasciatiis have 

 long filaments, like those of the 

 garfish, which entangle them in 

 masses or suspend them to various 

 objects. The large number of 

 fishes which carry their ova and 

 larva in their bucco-pharvngeal 

 cavity is a prominent feature, "and 

 Mr. Boulenger has found that it 

 is almost invariably the female 

 which does so, not the male, as in 

 such forms as .\rius. In some 

 cases (Haplochromis strigigena) 

 the male makes a small cavitv in 

 the sand where the eggs are fer- 

 tilised, the female thereafter taking 

 them into her mouth, and fasting 

 IVir a fortnight. Yet animals 

 much lower in the scale than fishes 

 do almost the same thing, as in 

 the case of Asterias Mullen, the 

 fertilised eggs and larvae of which 

 "-' '■ ' - -^■■- are borne in a mass by the 



parent over the mouth. The 

 Egvptian fishermen, however, explain the presence 

 of the ova in the mouth of the fishes very simply, 

 viz. by a "reversed method of parturition." The 

 whole subject, from the development of the nup- 

 tial tubercles in the males to the post-larval stages 

 of these remarkable Nile fishes, bristles with features 

 of interest. In addition, the field of fish-physiology is 

 inviting. Why is it that Polyptcrus bichir (a fish 

 which dies in tolerably fresh water if prevented from 

 reaching the air) cannot live in brackish water, and 

 that slight salinity kills it? whilst one species of fish 

 in Lake Menzaleh thrives either in fresh or salt water, 

 and another dwells equally in a hot spring at Makulla, 

 in the Persian Gulf, and in salt water all round the 

 Red Sea. The author takes in hand the explanation 

 of the peculiar coloration of Synodontis batensoda, in 

 which the ventral aspect of the body is darker than 

 the upper, viz. as a provision in connection with the 

 habit of swimming in a reversed position. Yet this 

 explanation will not avail for the post-larval Cal- 

 lionymus. -Again, are the habits of -Anabas in Africa 

 similar to those in India? 



