iI42 



a variable form as this the dimensions of the shell are by no means constant, 

 and cannot be considered as of any value for classification. Of course 



■ GossELET himself observes that his divisions are not varieties in the ordinary 

 sense, but rather groups of forms. He accepts that " un mejne individu 

 passe avec I'age d'un dans I'autre." 



Thus it seems rather better to the writer to adopt Peixizzari's rnethod 

 of classification for the Chinese examples than that of Gosselet, for the 

 material of the former writer all originated in China. Moreover, Peixizzari's 

 categories furnish sufficient subdivisio is for the fossils at the disposal of the 



• present writer. 



Pellizzari's classification on the other hand is by no means a compre- 

 hensive one, if one tries to apply it to a much larger material, for instance, 

 that of Gosselet. Most of the specimens of Gosselet may be grouped in one 

 or the other of Pellizzari's categories, but there will be some left which will 

 constitute transitional forms between some of them. Yet as far as the Chinese 

 material is concerned the division proposed by Peujzzari holds good at least 

 - for the present. 



Gosselet's divisions are as follows ; 



1) Cyliiidrici: Ratio of width to length greater that 3 " 



2) Attcmiati: „ „ 3 — 2.5 



3) lllongati: „ „ 2.5—2 



4) Hcniicycli: ,, ,, 2 — 1.60 



5) Proqiiadrati : less than 1.60 



6) ^Obovati: ?> 1.70 and the great- 



est width beneath the hinge- 

 line. 



The writer here ventures to reproduce the key to the classification pro- 

 posed by Pellizzari, which runs as follows. 



I. Cardinal line subequal or inferior to the maximum breadth of het 

 shell. Outline of the typical form lyratp (or approximately rectangular, 

 « quadrate or semicircular). 



.. A. Shell more or less depressed, or also swelled but not globular. 



