EFFECTS OF IISTBREEDHSTG AND CROSSBREEDING, 5 



that AC is markedly superior in the average for any given size of 

 litter. The latter fact is brought out in the index, obtained as in 

 the cases of the mortality percentages, by assigning weights of 1, 

 3, 3, and 1 to litters of 1, 2, 3, and 4, respectively. 



The differences among the experiments in the percentage of the 

 young which die before weaning have an effect on their standing in 

 average birth weight, since such young are in general below the 

 average. The average birth weight of the young which reach 33 

 dsLjs (Table 19) thus gives a better measure of the normal prenatal 

 growth rate and also is needed in calculating the rate of gain between 

 birth and weaning. 



The averages and indices for weight at weaning (33 days) are dealt 

 with in Table 21. The rate of gain between birth and weaning, 

 obtained from the data m Tables 19 and 21, is given in Table 20. 



ALLOWANCE FOR SEASONAL FLUCTUATIONS. 



Size of litter is of course not the only factor for which allowance 

 must be made in studying the effects of breeding on the various 

 characters. The most important factor is undoubtedly the environ- 

 mental situation. It would be highly desirable to compare the 

 records only in experiments which were conducted simultaneously 

 and under the same conditions. There are, however, practical diffi- 

 culties in carrying through such a project. Some of the present 

 experiments were carried on through the whole four years 1916-1919, 

 but others were started or dropped at intermediate times. Table 

 22 shows the average number of mature matings in each experiment 

 during each 3-month period, beginning with January to March, 1916, 

 and ending October to December, 1919. The number of litters, the 

 number of young, the number of young born alive, and the num- 

 ber raised in each experiment during each of these periods are shown 

 in Tables 23 to 26. 



The environmental conditions were unfortunately exceptionally 

 varied during the four years which we are considering. There were 

 three times — the winters of 1915-16 and 1916-17 and the spring of 

 1918 — when the stock did very poorly, due partly to unusual mnter 

 conditions, extreme variations m temperature, poor ventilation, etc., 

 but probably in the main to insufficient gi-een feed. In all of these 

 periods symptoms, such as lameness and bleeding at the gums, 

 appeared, which were probably indications of scurvy. There have 

 been no real epidemics of contagious disease, although many old 

 animals died in the periods noted above. The stock reached fairly 

 good condition during each fall; and good condition was maintauaed 

 through all of 1919. 



The records of the total inbred stock have been tabulated for each 

 3-month period during 1916 to 1919. A summary of the indices and 



