Manchester Memoirs, Vol. Ixiv. (1920) No. 2 3 



attenuatus and multiplicatus are derived from some common form 

 similar to but somewhat more primitive than multiplicatus. 



While the conclusions drawn by Mook are perhaps valid, the 

 more purely statistical portion of the work calls for certain 

 criticisms. Only 74 specimens of the form attenuatus were 

 measured; the curves for multiplicatus are based on the still 

 smaller number of 29 specimens. Yet in each case the percentage 

 of shells within any given range of index is returned to the nearest 

 one-tenth of one per cent., i.e., in the case of multiplicatus the 

 tacit assumption is made that the relations found for 29 specimens 

 would hold exactly for one thousand. A much more serious point 

 is that we are not told how the neanic measurements were ob- 

 tained. Mook appears to have regarded all his examples as 

 adult, however small they may have been; hence we find the 

 neanic measurements of various individuals exactly the same as 

 the adult measurements of other individuals; the latter therefore 

 would appear to be immature {e.g., neanic specimen no. 53=adult 

 no. 38, etc.). Again, the neanic width is always given as equal 

 to the adult width ; thus, it is assumed that after the neanic stage 

 has been passed the shell grows longer but not wider, the length 

 of the hinge-line remaining unaltered. While the growth-lines 

 are certainly much more crowded towards the ends of the hinge- 

 line than they are near the anterior margin of the shell, increase 



in length is always accompanied by increase in width, i.e., although 



dw . • 



diminishes very considerably as the shell increases in size, 



dl 

 yet it is always appreciable. For these reasons the neanic' curves 



of Mook are of very little value. 



The measurements of every specimen are published, and these 

 the writer proposes to use in order to obtain " growth-curves," so 

 far as the data will allow. 



In a recent note (4) the writer showed that in the case of the 

 recent freshwater bivalve Unio pictormn, Linne, the ontogenetic 

 curve of growth for the ratio antero-posterior length/dorso-ventral 

 length, as determined by measurements from the growth-lines 

 of individual shells, was in complete agreement with the phylo- 

 genetic curve obtained by measuring a large series of shells in 

 all stages of growth and determining the mean value of the 

 major axis which corresponded with each given dorso- ventral 

 length. It is proposed to investigate the growth of various 

 Brachiopoda by the method employed for the Unio above men- 

 tioned. 



