July 6, 191 1] 



NATURE 



well as Mycoderma vini and cerevisiae. A short 

 account is given of the history of fermentation and 

 of the controversy respecting spontaneous generation. 

 Then follow discussions of the biological relationships 

 of yeasts, variations in the Saccharomycetes, 

 morphology and anatomy of yeast-cells, spore-forma- 

 tion and its application to the analysis of yeasts, and 

 a number of allied subjects. ' Lastly come descriptions 

 of the more important culture and wild yeasts met 

 with in the brewing, distilling, and wine-making 

 industries. 



The last chapter gives a brief account of the methods 

 and apparatus employed in the preparation and trans- 

 port of cultures of pure yeast for industrial purposes. 



The translation has been on the whole well done, 

 although in some cases the English is stiff and the 

 grammar faulty. Rather ugly split infinitives occur 

 in moderately large number, and subject and verb 

 do not always agree in number. Use of the expression 

 "equal molecules of dextrose and lasvulose " is difficult 

 to defend, and "sorbite," "mannite," and "albu- 

 minoid" are nowadays better termed "sorbitol," 

 "mannitol," and "protein." Very few misprints are 

 noticeable ; d-methylglucoside (p. 359) obviously refers 

 to a-methylglucoside. T. H. P. 



THE FISHES OF AFRICA. 

 Catalogue of the Fresh-Water Fislies of Africa in the 

 British Museum (Natural History). By G. A. 

 Boulenger, F.R.S. Vol. ii. Pp. xii + 529. (Lon- 

 don : British Museum (Natural History), 191 1.) 

 Price 2I. 5s. 



THE author is to be heartily congratulated on the 

 appearance of the second volume of this great 

 work, which succeeds its predecessor after an interval 

 of two years, a period by no means excessive when 

 the amount of labour involved in a task of this nature 

 is taken into consideration. The present volume 

 completes the account of the carp tribe (Cyprinidae), 

 containing the great bulk of that group, and likewise 

 includes the whole of the cat-fishes (Siluridae), several 

 new genera and species being named. 



Although no one regards systematic works of this 

 class as the final aim and end of zoological science, 

 their importance and value cannot be overestimated, 

 since it is upon such sure foundations that all super- 

 structures of a more far-reaching and philosophical 

 nature must be based. That it was high time the 

 task of bringing our knowledge of the African 

 Cyprinidae and Siluridae up to date was undertaken will 

 be evident by a comparison of the number of species of 

 certain groups recorded in the present volume with 

 that given in Dr. Giinther's "Study of Fishes," pub- 

 lished in 1S80, and based on articles in the ninth 

 edition of the "Encyclopaedia Britannica." It is stated, 

 for instances, in p. 573 of the "Study of Fishes" that 

 the total number of species of cat-fishes of the exclu- 

 sively African genus Synodontis is fifteen, whereas Mr. 

 Boulenger has swelled the list to fifty-seven. Again, Dr. 

 Gunther estimated the total number of representatives 

 of the Old World cyprinoid genus Barbus at about 

 200, while Mr. Boulenger gives a list (inclusive of 

 nine additional uncatalogued species) of no fewer than 



xo. 2175, VOL. 8;] 



194 African members of the genus. Unfortunately 

 comparisons cannot be extended to the total numbers 

 of African Cyprinidae and Siluridae recorded by the 

 two naturalists, as Dr. Gunther enumerates only those 

 inhabiting Africa south of the Sahara; but, even so, 

 his lists of fifty-two Cyprinidae and sixty-one Siluridae 

 (op. cit. p. 230) inhabiting Ethiopian Africa are alto- 

 gether outclassed by Mr. Boulenger. 



This great increase in the number of African repre- 

 sentatives of the two families has, of course, a most 

 important bearing on previous conclusions as to the 

 place of origin of the two groups. Dr. Gunther 

 (op. cit. p. 225) suggested that since the majority of 

 the groups of fresh-water fishes common to India and 

 Africa, with the exception of the siluroid Clarias and 

 its relatives, had more representatives in the former 

 than in the latter area, the presumption is that they 

 are of Asiatic origin. Although these conditions are 

 now in many cases reversed, the conclusion will, we 

 think, still hold good in the case of Barbus, the mem- 

 bers of which, like many groups of antelopes with 

 ancestral forms in India, would appear to have under- 

 gone an unparalleled development when they reached 

 Africa. On the other hand, the abundance of silur- 

 oids in the Eocene of the Fayum, where no remains 

 of cvprinoids have hitherto been discovered, points to 

 the conclusion that this group is endemically African. 

 And here it is noteworthy that the connection between 

 the African and South American cat-fishes is now 

 regarded as even less intimate than was the case when 

 Dr. Giinther's work was written, the two African 

 species there referred (p. 233) to the South American 

 Pimelodus now being assigned to a genus apart. But 

 to pursue this interesting subject would demand more 

 space than can be given to it in these columns. 



If such space were available, we might presume to 

 criticise some details in Mr. Boulenger's "keys"; and 

 we cannot conclude without mentioning that the value 

 of the work would have been increased if the dates 

 of presentation of the specimens were added. The 

 work appears sintjularlv free from misprints. 



R. L. 



THE INTERNAL-COMBUSTION ENGINE. 

 Gas Engines. By W. J. Marshall and Captain H. R. 

 Sankey. Pp. xvi + 278. (London : Constable and 

 Co., Ltd., 191 1.) Price 6s. net. 



THIS book is the latest addition to Messrs. Con- 

 stable and Co.'s Westminster series, and is 

 intended to be useful to 



"those who, being either purchasers or users of gas 

 engines, wish to know the principles underlying the 

 design or construction and the methods of diagnosing 

 defects when they occur, and the steps to be taken to 

 remedy such defects." 



It may fairly be said to have achieved its purpose, 

 and, for the most part, any criticism to which it lays 

 itself open is little more than that to which almost 

 any first edition is liable. 



The book (an unusually heavy one to handle for 

 its size) is divided into ten chapters, of which the 

 first three deal with the theory of gas engines and 

 with the Otto and two-stroke cycles of operation. The 

 fourth, fifth, and sixth chapters are concerned with 



