NATURAL SCIENCE: 



A Monthly Review of Scientific Progress. 



No. 28. Vol IV. JUNE. 1894. 



NOTES AND COMMENTS. 



Composite Generic Fundamenta. 



'I'^HE American Geologist for April contains a note by Mr. J. M. Clarke, 

 ^ which is of such interest in connection with subjects lately 

 discussed in our own pages, that we take the liberty of quoting it at 

 length : — 



" Some remarks on ' Mathematical Biology ' in the February 

 number of Natural Science, and the suggestion made by Bather in his 

 recent work on the ' Crinoidea of Gotland,' that Galton's method of 

 composite portraiture might be employed to ascertain a specific type 

 from which variations in various degrees could be reckoned, recalled 

 to me some germane experiments in which I was interested some 

 eight years ago, and which indicate that there is even a possibility 

 among narrowly restricted, up-to-date genera, of educing, by a series 

 of comparative measurements and successive overlays, a sort of 

 generic fnndamentum. I take the liberty of citing and illustrating a 

 single instance of this among the fossils, the case being a simple one 

 and involving the search for the standard in one element of variation 

 only, namely, outline; but whatever may be done with a single specific 

 character, in this manner, is equally possible of all the others. 



" In volume v., pt. i, of the Palaeontology of New York are 

 described 57 species of the lamellibranch genus Leptodesma, two of which 

 are ascribed to the horizon of the Hamilton group, and 55 to the 

 fauna of the overlying Chemung beds. Leptodesma is an aviculoid 

 genus with smooth and convex valves, and the shells upon which 

 these observations are based are preserved plump and without 

 disturbance of the marginal outline. The genus in some of the 

 higher beds of the Chemung group is enormously prolific, so 

 that the perception and identification of these 55 specific values is 

 a puzzle to vv^hich the identification of Jurassic Ammonites or recent 

 Unios is but play. Nevertheless, no more instructive series of 

 variations of a single generic type has ever been described among 



2 D 



