580 ^- J- Crozier, 



dimensions include a double thickness of the thin periostracum, which 

 ammounted to 0,01 cm. 



For the right-hand valves the collected measurements are shown 

 graphically in Fig. D. Length is plotted on the axis of abscissas 

 and the other measurements on the axis of ordinates, the relative 

 magnitude of the several dimensions being such as to permit the 

 plotting of the curves without mutual interference. Many of the 

 points do not show on the plot, as they frequently fall one upon 

 another, some circlets therefore representing duplicates or even 

 triplicates, etc. The figures for the left valves give identical results, 

 and are therefore not reproduced. 



Discussion. 



It is noticeable that the internal characters (interadductor 

 distance and palliai sinus depth) are distinctly more variable than 

 the depth and width of the valves, as expressed in terms of the 

 valve length. The correlation between the measured characters is 

 close. The smooth curves best satisfying these observations are 

 straight lines. Taking the origin at (0,0) the relations of the 

 other dimensions to the maximum length of the valve (L.) may be 

 expressed by equations of the form 



y = a x + b 

 where x = length 



y = depth, width, etc. 

 and a gives the rate of growth of the dimension concerned with 

 respect to the growth of the valve length. 



The values calculated for these curves are 



Depth D = 0,912 L + 0,025 



Width w = 0,149 L + 0,070 



Interradductor distance I = 0,517 L + 0,14 



Palliai sinus depth P = 0,343 L -j- 0,26 



Now, the width of the right valve being found equal to that 

 of the left, we may write 



2 w = W = 0,30 L + 0,14 



Inasmuch as the curves for Depth and Width pass fairly closely 

 through origin, some of the above facts may be collected into one 

 formula : 



