PARKER AND BULLARD. — LITTERS AND NIPPLES IN SWINE. 409 



interpretation and we are inclined to regard the signs of segmentation 

 as of quite secondary origin, perhaps the result of the segmental 

 tendencies of the trunk exerting themselves on what would have been 

 otherwise an irregular series. From this point of view we regard the 

 irregular arrangement as the more primitive and the regular one as of 

 later origin. Physiologically the best arrangement for the nipples 

 would probably be one of alternation in the two rows such as is 

 partially met with in what Wentworth has called the triangular 

 pattern, for this would admit of the greatest freedom of access to the 

 nipples for the young pigs as the sow lies on her side. Such an ar- 

 rangement of the whole series of nipples is extremely rare, however, 

 though what we have called the irregular plan approaches this more 

 nearly than the regular one. 



Relation of the Size of the Litters to the Number of Nipples 

 IN Individual Pigs. 



We have now gathered the necessary data for a comparison of the 

 size of the litter with the number of nipples in individual pigs. These 

 relations are shown in Table 13 in which the sizes of the litters are 

 plotted against the numbers of nipples in the females. 



The number of pigs in a litter studied by us ranged from 1 to 15 

 with 6 as the mode. The number of nipples in female pigs ranged 

 from 8 to 18 with a mode of 12. These figures show at once that no 

 simple and close correlation is to be expected between the two series 

 of records and this opinion is confirmed by the coefficient of correlation 

 calculated from Table 13, namely 0.0035 ± .0124. This lack of cor- 

 relation is also represented in the last column in Table 13, in which is 

 given the mean number of nipples for litters of different sizes. An 

 inspection of this column shows at once that there is no obvious 

 tendency for the number of nipples to vary with the size of the litters ; 

 in other words, in swine, as in sheep (Bell, 1904, p. 768), the number of 

 nipples is not obviously related to fecundity. 



The relation of the size of the litters to the number of nipples can 

 also be shown in superimposed frequency polygons, Table 14. Al- 

 though there may be as few nipples as 8 and as many pigs in a litter 

 as 15, it is quite evident from the nature of the overlapping of the 

 polygons that disadvantageous combinations in which the number 

 of young pigs outrun the provision for milk, cannot be of frequent 

 occurrence and that in general, as indicated by the mode of 6 for the 

 litters and 12 for the nipples, there are about twice as many nipples as 

 young. 



