426 



SCIENCE. 



[N. S. Vol. XIX. No. 480. 



both * epi- ' and ' basi- ' and ' cerato-branchial ' 

 divisions, and the ' quasi-maxillary ' cartilage. 

 For if these structures are present, our knowl- 

 edge of Palceospondylus has made a forward 

 step and an important one. For it excludes 

 one hypothesis, that of a Devonian lamprey. 

 I confess, however, much as I am in sympathy 

 with this result, that a critical examination of 

 the present plates does not convince me that 

 the authors have carried their point. When 

 one bears in mind the fact that the head re- 

 gion of the tiny fossil is flattened out of rea- 

 son, bitumenized, with parts displaced, with 

 irregular contours now separate, now con- 

 fluent, it is difiicult to see, for example, why 

 the ' maxillaries ' should be anything more 

 than the rim of the ' hemidome ' (cf. Figs. 

 10, 11, 12), or the most conspicuous 'gill 

 arches ' more than the anterior and posterior 

 rims of the ' otic ' mass. And there is in 

 fact variation enough in these elements, even 

 in the figures given, to warrant our skep- 

 ticism. Indeed, if one has still any faith in 

 the preservation of intricate interrelationships 

 of delicate elements in Palceospondylus, he 

 has only, I believe, to consider the shapeless 

 condition of the neighboring vertebral column 

 and fin supports as shown in any of the models, 

 for we are morally sure that centra and fin 

 rays never existed in the living animal in such 

 an amorphous mass as here represented. 



Grateful we certainly are to these pains- 

 taking authors, for the facts they present are 

 desirable, even though we may be disappointed 

 in their content. But the chiefest virtue of 

 their research is to my mind this — that by the 

 micro-section method we can add little of im- 

 portance to our knowledge of this form. For 

 in spite of siich a method, and admirably car- 

 ried out, no certain details have been added to 

 those already detected in the usual way by the 

 keen eye of Dr. Traquair. And if this is true 

 one need hardly add that what is especially 

 needed in our future dealings with Palceo- 

 spondylus is less memoir and more material. 

 In this regard I can not repress the belief that 

 the paleontologist who will spend some time, 

 possibly months, in the little quarry at Achan- 

 arras will yet solve the puzzle. There is cer- 

 tainly evidence that although five out of ten 



of the fossils measure between 20 and 30 mm. 

 some certainly occur which are over 50 mm. 

 in length, and that between the largest and 

 the smallest there are gradations in the pro- 

 portions of head and column both in length 

 and thickness. And if this be true, why may 

 not Palceospondylus prove a larval form, and 

 in this event more abundant material might 

 reasonably prove what the adult is 'i The latest 

 authors ' far from deny that some change in 

 the proportional size of the organs of 

 Palceospondylus has taken place with growth,' 

 but I think they would have been more im- 

 pressed had they examined better preserved 

 specimens, and those especially with a greater 

 range in size. Their smallest specimen prob- 

 ably measured little under 18 mm., for it 

 measured 14 mm. and lacked obviously the 

 tail tip. On the latest evidence, therefore, we 

 can hardly deny the possibility that Palceospon- 

 dylus was a larval form. Bashford Dean. 

 Columbia Univeesitt. 



Catalogue of Keyboard Musical Instruments 

 in the Groshy Brown Collection, 1903. The 

 Metropolitan Museum of Art. New York. 

 4to. Pp. 313. Price, $1.00 ; express, 25 cents. 

 In continuation of the series of catalogues 

 already noticed in Science (N. S., XV., p. 943, 

 1902) the present sumptuous volume has re- 

 cently been published. 



In 127 half-tone plates made from photo- 

 graphs from the originals, 84 keyboard instru- 

 ments are shown. No such collection of 

 instruments is found elsewhere in the world, 

 and no sxich collection of illustrations is avail- 

 able in any other book, or dozen books. There 

 are 43 plates of 24 plucked instruments — 

 psaltery, spinet and harpsichord; 47 plates 

 of the 32 instruments with struck strings — 

 dulcimer, clavichord and piano ; 2 plates of 

 bowed instruments; 28 plates of the 21 key- 

 board wind instruments — regals and organs; 

 and 7 plates of 5 pianos with metal or glass 

 bars ; then follow 7 plates illustrating actions. 

 Many of them are full-page plates. There 

 is a brief description of each instrument, and 

 the late A. J. Hipkins (remembered by phys- 

 icists as associated with A. J. Ellis in some 

 of his important work on scales) has furnished 

 a valuable short introduction which points out 



1^1 



Mx:: 



