804 



SCIENCE 



[N. S. Vol. XXVIII. No. 727 



carapace or any part of it " fascia bones " ; 

 they must certainly have arisen as purely 

 dermal, membrane bones, and must have pre- 

 ceded the development of an outer layer. And 

 one can not understand why such ossicles 

 might not have developed, under the stress of 

 peculiar environmental conditions, in the 

 aquatic turtles after the loss of the true 

 dermal carapace. Furthermore, this phy- 

 logeny seems yet to be based almost wholly 

 on hypothesis, for we have little evidence of 

 such a primitive condition, save that possibly 

 afForded by the neural ossicles of Toxoclielys. 

 Here, too, it seems to the writer the argument 

 is against the hypothesis, since the more spe- 

 cialized aquatic Toxoclielys has the neural 

 ossicles, while the nearly related, and less 

 aquatic Porihoclielys is without them. 



The usual suborders, Cryptodira, Amphiche- 

 lydia, Pleurodira and Trionychoidea are made 

 superfamilies by the author, and, so far as the 

 Trionychoidea are concerned at least, the 

 writer agrees with him under any classifica- 

 tion. 



Dr. Hay recognizes the difiSculty in deriving 

 the turtles from any except the most primi- 

 tive of reptiles; in other words, the order 

 represents a phylum all its own in the evolu- 

 tion of the reptiles, a view first offered by 

 Cope, with which the present writer is wholly 

 in accord. 



From his wide and accurate acquaintance 

 with the literature of American fossil verte- 

 brates it would be expected that little has 

 escaped the author's attention; but he is not 

 infallible. He expressly states that no fossil 

 turtles are known from the Dakota Cre- 

 taceous, overlooking the fact that only a few 

 years ago (1899) a very interesting specimen 

 from that formation was described and figured 

 by Parmenter in the Transactions of the 

 Kansas Academy of Science; and it is to be 

 regretted that he did not examine and describe 

 the remains of a large marine turtle from 

 the Benton of Kansas now in the University 

 of Kansas Museum, to which attention was 

 called six years ago by the present writer in 

 the Kansas University Quarterly. However, 

 such omissions will be found to be very rare 

 in the work, and are not disquieting. 



The illustrations are for the most part good, 

 especially the text figures, of which there ar« 

 over seven hundred. The Carnegie Institu- 

 tion is to be congratulated on the publication 

 of this valuable and useful work. 



S. W. WlLUSTON 



The Study of Stellar Evolution. By Geo. E. 

 Hale. Pp. xi + 252; 104 plates; 7 text- 

 figures. University of Chicago Press. 

 1908. Price, postpaid, $4.27. 

 The past quarter of a century has witnessed 

 a complete revolution in the conception of 

 astronomy and of an astronomical observa- 

 tory. The astronomy of twenty-five yesirs 

 ago was the science of position and of motion. 

 The problems which then confronted the in- 

 vestigator were those concerning the size and 

 shape of the planets, their distances from the 

 sun, the periods of time in which they com- 

 plete their orbits, and the discovery and ex- 

 planation of the laws which govern their 

 motions. The solution of these problems re- 

 quires the precise measurement of the positions 

 of the various bodies in the heavens at fre- 

 quent intervals covering long periods of years. 

 The old observatories were built and equipped 

 with this end in view; the fundamental in- 

 struments were the meridian circle and the 

 clock. The chief use of the equatorial was to 

 measure the position of objects too faint to be 

 seen with the meridian circle. Photography 

 and spectrum analysis, as applied to astron- 

 omy, were beginning to obtain recognition, 

 but were hardly regarded as the work of a 

 '' simon-pure " astronomer. 



To-day this is changed; the astronomer is 

 concerned not so much with the position and 

 motions of the bodies, as with their physical 

 characteristics; he wishes to know what they 

 are, not where they are; from what they de- 

 veloped and what their future life history will 

 be, rather than the exact path through space 

 which they have traveled and are traveling. 

 To solve these problems an astronomer to-day 

 must be a physicist and a chemist, as well as 

 a mathematician. The simple observatory of 

 the past has become a great complex labora- 

 tory, in which the spectroscope of the phys- 



