54 



The Jol'rxal of Heredity 

 Weight of Mothers of Female Lambs {Continued). 



DIFFERENCE DISAPPEARS. 



"The results," as he notes "are 

 rather startHng. It would be unneces- 

 sary to give a graphical diagram, for 

 it is obvious from inspection of the 

 tables that the average weights of the 

 mothers of male and female lambs 

 taken at the end of October, November 

 and December, of the year preceding 

 the birth of the lambs were the same. 

 So that the two curves would be abso- 

 lutely identical. 



"What this means I do not yet know. 

 The first im])ression made upon the 

 mind is, that the former results were 

 fallacious because based upon small 

 numbers. In the above case the smallest 

 nimiber of cases used in determining an 

 average is 100, so that the averages are 

 surely more reliable than any yet 

 obtained. It may be possible that 

 significant variations in weight may 

 occur within a period of two or three 

 weeks, without affecting the monthly 

 averages. If these significant variations 

 occur chiefly in the month of October, 

 this might account for the lower average 

 weight of the male-bearing ewes at the 

 end of Sei)temljer. The only difference 

 that I can oljserve is this, that the male- 

 bearing ewes increased in weight at a 

 greater rate than the female-bearing 

 ewes during the month of October. 

 They started lower and reached the 

 same point by the end of the month. 



The difference, however, is so slight that 

 it may not be significant. 



"The only way to ascertain whether 

 there is any real difference in the average 

 weights of the two classes of ewes 

 (male-bearing and female-bearing) with- 

 in a short period of time, two or three 

 weeks, is to multiply observations at 

 short intervals of time." 



Having found the apparent result of 

 the first calculation on determination of 

 sex nullified by the addition of further 

 data. Dr. Bell decided not to press the 

 investigation of the influence of the 

 same factor — special feeding — on the 

 fecundity of the ewes, until he had 

 larger figures on which to base his com- 

 putations. On this point he noted: 



"In order to examine the results as 

 to twins I added together the weights 

 of the mothers of lambs born during the 

 decade 1890-1899, the year 1900, and 

 the year 1901. A marked difference 

 exists between the weight-curves for the 

 two classes of ewes, twin-bearing, and 

 single-bearing. In the case of the twin- 

 bearing, however, the absolute numbers 

 involved are too small to make me 

 certain of my results, especially after 

 the ex])erience with the male and 

 femalc-])earing. The maximum number 

 of cases used in determining an average 

 is only 88, and the minimum 50 cases. 

 It would be well in this case also to have 

 the average of at least 100 cases before 

 announcing results." 



HANDLING THE MULTI-NIPPLED SHEEP 



Dr. Bell's final decision to pul)lish 

 the results of the investigation was due 

 partly to the l)elicf that his own ex])eri- 

 ence might ser\-e as an example 



of the danger of drawing conclu- 

 sions in such studies from anything 

 l)ut a large mass of data. The report 

 has high value, liowever, merely as a 



