February 27, 1902] 



NA TURE 



389 



to the regrettable fact that the author has been obliged 

 to abandon the use of a very important character in 

 the definition of the higher groups, viz. the presence or 

 absence of the inesocoracoid or " precoracoid" arch, 

 the presence of which in the Isospondyli as under- 

 stood by Cope had been duly emphasised in the 

 diagnosis of this suborder at the end of the preceding 

 volume. The Isospondyli are now made to include the 

 Haplomi, an innovation with which the reviewer is 

 unable to agree for the reason that the study of recent 

 fishes proves their separation to be absolutely necessary. 

 Even where the character of the " precoracoid process " 

 is appealed to for the definition of families, error has 

 crept in, at least in the case of theGonorhynchids, which 

 are stated to be devoid of it, and hence are unjustifiably 

 regarded as " only slightly modified Scopelids." Other 

 points in the identification of the elements of the pec- 

 toral arch seem in need of revision, as in the Thrisso- 

 pater figured on plate vii., where the bone named 

 " postclavicle " either represents the supraclavicle or 

 overlies the latter and the clavicle, in which latter case 

 it does not, as I believe, answer to the definition of the 

 ElopidK, and in the restoration of Eurypholis boissieri, 

 p. 206, where the basalia of the pectoral fin are repre- 

 sented as attached to a bone termed " postclavicle." In 

 the definition of the Apodes, " pectoral fin with more 

 than five basalia " is true of AnguiUa, but does not apply 

 to Conger and other genera. 



It is clearly often impossible to assign e-xtinct fish- 

 remains to their systematic position with that rigid pre- 

 cision which may be attempted in the case of living 

 forms. Dr. Woodward, as he tells us in the introduc- 

 tion, has therefore deemed it advisable to adopt a 

 broad conception of families and genera more in accord- 

 ance with that of Dr. (iunther than with that of later 

 writers. But his classification, on the whole, is greatly 

 ahead of that followed in Zittel's manual and in the 

 text-books published in this country. He has amply 

 availed himself of the reforms introduced by Cope and 

 by Sagemehl. The arrangement of the great group of 

 Acanthopterygians is still the most unsatisfactory, the 

 definition of its subdivisions being of a very provisional 

 nature and lacking in precision ; groups like the Bery- 

 ciformes, Cha-todontiformes and Blenniiformes are cer- 

 tainly quite artificial, and the new sense in which these 

 terms are used must be regarded as a retrograde step. 

 .Some explanation might have been given by the author 

 of the reasons that have induced him to place the 

 Blochiid:c among the Blenniiformes rather than among 

 the Scombriformes. 



The fossil forms dealt with under the Isospondyli 

 offer a highly interesting and suggestive gradation from 

 the later Ganoids to the earlier .Acanthopterygians, such as 

 the BerycidiP, so abundant in Cretaceous formations, but 

 we are unfortunately still without a clue to the derivation 

 of the eels proper, or Apodes, degenerate fishes which are 

 traced back to the Chalk. Among these, Urenchelys, 

 from the Chalk of Mount Lebanon, is shown to differ 

 from existing genera of the same family in having a small 

 caudal fin supported by expanded hypurals, thus showing 

 the "diphycercal" condition which prevails at the present 

 time to have been derived from a " homocercal." The 

 Percesocine genus Cobitopsis settles once for all the vexed 

 NO. 1687, VOL. 65] 



question of the systematic position of our sand-launce, 

 Ammodytes, as it has retained the abdominal pelvic fins 

 which have entirely disappeared in the existing genus ; 

 Ammodytes must hence be removed from the Ophidiid 

 Anacanthines and placed near the Scombresocids or 

 gar-pike and allies. 



The publication of the " British Museum Catalogue of 

 Fossil Fishes " marks a great advance in ichthyology, 

 and we heartily congratulate Dr. Smith Woodward on its 

 completion. It is announced in the preface that the 

 author proposes to prepare, in the course of the present 

 year, a supplement giving a list of additional important 

 genera discovered and published since the earlier volumes 

 were issued, the first dating as far back as 1889 ; also a 

 stratigraphical table showing the appearance in time of 

 families and genera of fossil fishes, together with a 

 general index to the four volumes. 



A last word as to the illustrations. The plates, as well 

 as the outline figures in the text, are excellent, both from 

 the point of view of artistic finish and scientific accuracy, 

 and do the greatest credit to the artist, Miss G. M. 

 Woodward. The intercalation of a collotype plate (xvii.) 

 is, however, to be regretted, as not in keeping with the 

 style of the other illustrations and quite superfluous, the 

 figures having already appeared elsewhere, although no 

 allusion to this is made in the accompanying explanation. 



G. A. B. 



OUR BOOK SHELF. 



Tales of a Dying Race. By Alfred A. Grace. Pp. x-(-25o. 



(London : Chatto and Windus, 1901.) 

 The title of this little book is somewhat misleading. 

 Out of eight-and-twenty tales, only four are, properly 

 speaking, Maori tales. The rest are stories of the 

 contact between the Maories and the white settlers, 

 traders and missionaries. Even the four Maori tales are 

 retold \X\ pa/ccha fashion, until there is little of the Maori 

 left in them beyond the skeleton. The majority have 

 already appeared in antipodean periodicals. They are 

 all charmingly told, and, illustrating as they do many 

 sides of the Maori character and the romance of earlier 

 days of the colony, they form a worthy tribute to the 

 noblest of savages, and cannot fail to rouse vivid feelings 

 of regret that the race is doomed to extinction. Mr. 

 Grace writes of the people and their surroundings with 

 keen sympathy, the full secret of which is not disclosed 

 until the last story, in which he relates an adventure of 

 his early life as a missionary's son, when his mother and 

 her children were rescued from an impending and 

 horrible death by the unflinching courage and fidelity of 

 a native chief He has done well to preserve the narra- 

 tive, as well as the other contents of this entertaining 

 book, in a permanent form ; but he himself would hardly 

 claim scientific value for the collection. 



Lehrbitch dcr Diffcrentialglcichu7!gen. \'on Dr. H. 



Liebmann. Pp. vi -I- 226. (Leipzig : Veil and Co., 



1901.) 

 Thls interesting and well- written book shows that the 

 ideas of Sophus Lie are at last bearing fruit, even in 

 elementary text-books. There are three chapters deal- 

 ing respectively with ordinary differential equations of 

 the first order, with similar equations of higher order 

 and systems of such equations, and with panial differen- 

 tial equations of the first order with two uidependent 

 variables. Besides this, there is an introductory chapter 

 dealing mainly with existence-theorems, and a concluding 

 one on partial differential equations of the second order. 



