450 Mr. W. F. K. Weldon. The Variations [Apr. 17 y 



The following table will show the relation between this curve and 

 the normal curve of error : — 



Table II. — Curve of Distribution of Carapace Lengths — Plymouth. 



Grade. 



Ordinate. 



Ord.-M. 



urct. — OIL 



Q. 



Normal curve. 



5° 



261 -50 



+ 10 -98 



+ 2*42 



+ 2-44 



10 





4- 8 '43 



T O tO 



4- 1 *8f5 



T J. OU 



4-1 'QO 



T X (J\J 



20 



256 -05 



+ 5 53 



+ 1-22 



+ 1-25 



25 



255-07 



+ 4-55 



+ 1-00 



+ 1-00 



30 



254-10 



+ 3*58 



+ 0-79 



+ 0-78 



40 



252 -27 



+ 1-75 



+ 0-39 



+ 0-38 



50 



250 -52 



o-oo 



0-00 



o-oo 



60 



249 -09 



- 1-43 



-0-32 



-0-38 



70 



247 -29 



- 3-23 



-0-71 



-0-78 



75 



246 -00 



- 4-52 



-1-00 



-1-00 



80 



244 -74 



- 5-78 



-1-28 



-1-25 



90 



241 -39 



- 9-13 



-2*10 



-1-90 



95 



238 -60 



-11 -92 



-2-63 



-2-44 



It will be seen therefore that the average length of the carapace 

 in the Plymouth specimens was 250*52-thousandths of the body 

 length, and that the frequency with which this length and the 

 various observed deviations from it occurred was almost exactly that 

 indicated by a curve of error whose prob. error = 4*53. 



In the above table, all the steps in the determination of the co- 

 efficients of Q are indicated. To indicate all the steps in this deter- 

 mination in the case of each race would involve a great deal of vain 

 repetition; and, therefore, in the following table the coefficients 

 themselves are alone indicated. It will be understood that the entry 

 opposite each grade in this table is found by subtracting the value of 

 M from the observed ordinate at that grade, and dividing the 

 remainder by 2(Qi~~Q3)* ^ ne value thus obtained should be the 

 coefficient of Q in the table of ordinates of a normal curve given on 

 page 448. These coefficients are repeated in the last column of the 

 following table. 



