Table 42 . — Frequency of antagonistic encounters according to the locality of the place of birth of the participants . 
Sex of the winner listed first ( expected frequency in parenthesis) 
Sex of the 
pair 
WINNER 
Born in 
southeast 
half of pen 
Born in 
northwest 
half of pen 
Born in southeast half of pen . 
LOSER 
Born in northwest half of pen . 
Male-Male . . . 
Female-Female 
Female-Male . . 
Total . . 
Male-Male. . . 
Female-Female 
Female-Male . . 
Total . . 
30 (23.9) 
4 ( 6. 2) 
8 ( 4. 5) 
42 (34.6) 
16 (11.7) 
14 ( 5.0) 
2 ( 2 . 2 ) 
32 (18. 9) 
5 (11.7) 
1 (5.0) 
1 ( 3. 6) 
7 (20.3) 
2 ( 5. 7) 
1 ( 3. 9) 
1 (1.7) 
4 (11.3) 
unless the number of sexually mature animals born 
in each half of the pen is taken into consideration. 
On the southeast side, 51 males and 34 females, 
who were born there, lived more than 150 days. 
Correspondingly, there were 25 males and 27 
females from the northwest side. This refers to 
individuals born in 1948. Inclusion of the 22 rats 
born during 1947 does not materially affect the 
proportions of sex to locality of birth. 
The hypothesis used for determining the ex- 
pected frequency of interactions was that they 
should be dependent upon the relative numbers 
of each sex and place of birth available for inter- 
action. Thus, if a denotes the number of southeast 
males and b the number of northwest males, the 
interactions would be proportional to a 2 (southeast 
with southeast) + 2ab (southeast with northwest) 
b 2 (northwest with northwest). Furthermore, the 
ab ' s on a purely random basis should be divided 
between southeast winners and northwest winners. 
The 53 observed male-male encounters would 
be (0.450<z 2 )+(0.441a6)+(0.108& 2 ). The expected 
female-female and female-male encounters were 
similarly derived. 
Certain general conclusions may be drawn from 
the table: (a) There were more antagonistic inter- 
actions than expected, when both members of 
the pair were rats of higher social status from the 
southeast half of the pen; while there were fewer 
encounters than expected when both members 
of the pair represented rats of lower social status 
from the northwest half of the pen (Chi Square 
6.30, P between 0.02 and 0.01). (b) Among the 
pairings involving one rat from each half of the 
pen there was a greater than expected number of 
winners representing the rats of higher social 
status from the southeast half of the pen (Chi 
Square 17.80, P less than 0.001). 
There is one question left unanswered in this 
study. This is: Do rats which engage in fewer 
antagonistic interactions also have fewer social rela- 
tions of all kinds? In this connection it is worth- 
while to mention the results of some studies {16) 
I conducted at the Roscoe B. Jackson Memorial 
Laboratory during 1950 and 1951. Two groups of 
free-ranging house mice were studied. One con- 
tained 8 males and the other 11. The status of the 
individual in every observed interaction was re- 
corded; whether it was dominant, subordinate, or 
neutral. In each colony the following relationship 
was exhibited. As the number of positive inter- 
actions (dominant or subordinate) per individual 
declined there was a decline in total interactions. 
In other words, in the absence of positive inter- 
actions there was not a corresponding increase in 
neutral interactions. Furthermore, those individ- 
uals with the most total interactions had the great- 
est percentage of these dominant. 
197 
