E. H. J. Schuster 



201 



Diagrain I. sIhjws thc regressioii lincs of Meusuromcnt No. 1 (in tlic carapace 

 for thesc twü dopths ; tho ilotttMl liiio is tor tlie deup water foniis, the ooiitiiiuoiis 

 lino for tlie .shallow watcr forins. It will bt; iiDÜced that one carapace Iciigth is 

 as.sociated wiih a iiiucli smallcr value of Measuruiiicut No. l in the dcep water 

 fonii.s thaii in tili' .shallow. 



Di.UiRAM I. ^[l^usu)■^•nu•llt Xo. I. 



Comparison between Regression Lines of Me;isui-ement No. I on Carapace Lengtli between Shallow 



Water <f<?(BB) and Deep Water ^<J(DD). 



Shallow Water ii. 



Mean Measurement No. 1, 10'2718 mm. 



Mean Carapace Length 8'40(i3 mm. 



Regression of Measurement No. 1 on Carapace Length 1-6'212. 



Deep Water ii. 

 Mean Measurement No. 1, 9-707H. 

 Mean Carapace Length 8-.58.54. 

 Regression of Measurement No. 1 on Carapace Length 1'5028. 



Table XI. shows thc correlation between the length of the right chela and the 

 length of the carapace for deep, Table XII. for the shallow water </ forms. 



For the deep water fornis r = -DoSO + 0036, 

 For the shallow water forms r = -9503 + '0028, 

 Ditfereuce = 0114 + 0046. 



Biometrika ii 26 



