On the grounds that equal facilities were avail- 
able on the northwest and southeast side of the pen 
for reproduction, and therefore, there should have 
been an equal reproduction on the two sides of the 
pen, there was a significantly (Chi square, 8.89, 
p )>0.01) greater number of rats weaned on the 
southeast side of the pen (table 29). However, 
there were no significant differences in mortality 
with references to place of birth. 
Even so, the proportionately greater mortality 
before May 1949 of the northwest born rats, who 
lived over 150 days, provides additional support 
to the belief that the northwest half of the pen 
developed into a less favorable place of living 
(middle column, table 29, Chi Square 3.60, p 
circa 0.06). 
Shifts in places of residence from the time of 
birth in 1948 to the places where the rats were 
found in May 1949 (table 30) succinctly reflect 
the sociological forces in operation. Of those rats 
born on the southeast side of the pen, twice as many 
males migrated to the socially more unfavorable 
northwest side of the pen as remained near their 
place of birth; whereas only half as many females 
made this shift as remained at their place of birth. 
This significant sexual difference in frequency of 
migration by members of the socially higher rank- 
ing groups of rats indicates the greater social 
pressure experienced by males. In contrast to this 
situation where over half of the members of the 
socially higher ranking groups of rats established 
themselves in the half of the pen inhabited by lower 
ranking rats, none of these lower ranking rats of 
the northwest side of the pen established them- 
selves in the southeast side (comparison b, table 
30). Thus, the separate phenomena of (a) the 
greater number weaned on the southeast side of 
the pen, (b) the greater survival rate of southeast 
rats, (c) the marked invasion of the northwest 
half of the pen by the southeast rats, and (d) 
the failure of any northwest rats to invade the 
southeast half of the pen, all contribute to a sig- 
nificant redistribution (comparison c, table 30) of 
the members of the population. 
One of the consequences of the social stratifica- 
tion and redistribution of the population was a 
differentiation in reproductive rates. At the time 
of the termination of the colony there were 21 
surviving litters. Many other litters conceived 
during the spring of 1949 /"see later section on 
stress and reproduction, pp. 214-216) were either 
resorbed or lost shortly following parturition. 
Therefore, when consideration is made of the 
marked strife that was existing within and between 
members of different groups of rats, it is unlikely 
that manv more litters would have survived in the 
absence of a marked drop in the adult population. 
In fact any such survival of rats after May 1949 
would probably have increased the discrepancy 
in reproductive rate between rats born in the 
southeast and northwest halves of the pen during 
1948 (see table 31). For the 44 surviving adult 
females born during 1948, 21 had surviving litters 
and 23 were without surviving litters. If the ratio 
of 0.477 with litters to 0.523 without litters is 
taken as the expected ratio, it may be seen that the 
females born on the southeast side of the pen, and 
who remained there, exhibited a significantly 
better survival of their litters. On the other hand 
the northwest females had a significantly poorer 
survival of their litters. The survival of litters 
born to southeast females, who migrated to the 
northwest half of the pen was poorer than those 
which did not migrate, but was better than that of 
the northwest females. This is a reflection of the 
more favorable conditions surrounding the early 
life of the southeast born females. 
This differential in reproductive performance, as 
fully shown in table 31, is particularly striking if 
we consider only those females who remained in 
the same half of the pen in which they were born. 
The ratio of successfully reproducing to unsuccess- 
fully reproducing rats was 14:4 for southeast fe- 
males and 2:14 for northeast females. The chi 
square for this difference is 14.335 with a prob- 
ability of chance occurrence of less than 0.001. 
Mention has already been made of the inference 
that such differential reproduction should lead to 
a greater degree of genetic homozygosity than one 
would anticipate for the same number of randomly 
breeding rats. Such a trend toward greater homo- 
zygosity should be further enhanced by the differ- 
ential sex ratio. In 1949 the ratio of males to fe- 
males was 11:18 for the southeast rats and 41:26 
for the northwest rats. The chi square for this 
difference is 4.397 with a probability of chance 
occurrence between 0.05 and 0.02. The effective 
breeding population may be roughly approximated 
by 0.5 of the total females and 0.25 of the total 
males, and furthermore, these breeding individuals 
tend to be more closely related than do their non- 
breeding associates. However, any trend toward 
more rapid development of genetic homogeneity 
produced by this restriction of the actual breeding 
142 
