EFFECTS OF IRRADIATION ON BEHAVIOR AND BLOOD 695 



differences between the 3 experimental groups were determined by analysis 

 of variance with respect to the preirradiation and postirradiation differences. 

 If the F-ratio obtained was not statistically significant, the data from all 

 experimental animals were combined and were compared with the scores of 

 the control group by means of a t-test. When a significant F-ratio was ob- 

 tained, the means of the separate groups were compared with the control 

 group by means of t-tests. 



Results 



Food and Water Intake 



Determined in somewhat different ways, the results of our two series are 

 essentially similar. Figure 2 shows the results for 300 and 400 r in the first 

 series. This study indicated that the threshold for anorexia lies somewhere 

 below 300 r and perhaps above 100 r. The second finding was unexpected, 

 because food intake was lowest on irradiation day and lower in the 1st week 

 than in subsequent weeks. 



In the second series, (Fig. 3), although the curves fall in a logical order, 

 statistically significant changes in food intake related to dosage level could 

 not be established with 4 animals in each dosage group. However, the 3 

 experimental groups collectively ate less than the control groups during the 

 1st postirradiation week and this difference was significant at the 1% level 

 of confidence. 



The early onset of anorexia noted in the first study is thus confirmed and 

 again the data suggest that the threshold for anorexia is below 300 r. 



The early changes are maximal. In fact, in the following 5 weekly periods, 

 food consumption of the combined experimental groups did not differ sig- 

 nificantly from that of the control group. There is, however, a second, 

 though less severe, reduction in food intake during the 3rd postirradiation 

 week that occurs in the groups exposed to 400 r and to 500 r. 



Water intake roughly paralleled food intake, being maximally reduced in 

 the 1st postirradiation week. The difference between the mean for the com- 

 bined experimental group and the control mean was significant at the 5% 

 level of confidence during the 1st week. The difference for the subsequent 

 weekly periods was not significant, but the decrease over the entire post- 

 irradiation period was significant at the 1% level of confidence. 



Activity 



In our first series of experiments, activity was recorded objectively by a 

 strain-gauge penwriter and subjectively by human raters. The objective 

 method indicated some reduction of activity in the first postirradiation day 



