514 MISS O. C. LODGE ON THE 



and this was with a 30 % formalin bait (i. e., 25 c.c. of 30 "/o 

 fonnalin in 75 c.c. of casein mixture), making 7'5 % of pure 

 formalin (Table III.). Amongst substances which did not kill 

 until the second day, excluding No. 5 * of Table IV., ammonium 

 nitrate {i. e., 5 7o P^^i'e NH^ISTOg) gave the highest percentage 

 (53 7o)- Next came 40 °/„ formalin (45 7o)> ^^^^ then antimony 

 oxychloride (37*5 7o)- ^or the percentages with the other 

 substances, see Table IV. 



Other experiments were also made to see what Mas the effect 

 of first feeding a poison-bait to the flies in the cylinders, and 

 then a clean bait to the survivors. Details of these experiments, 

 giving percentages of deaths, etc., will be found in Table V. 



The results seem to show that the flies died after a longer 

 or shorter time from the effects of feeding upon the poisons, and 

 not from starvation due to not feeding on the baits because they 

 were distasteful, for in most cases flies were seen to feed. Also, 

 in the experiments on starving flies, it was seen that these flies 

 could live longer before succumbing than did the flies which had 

 been given poison-baits ; for in these the percentage of deaths 

 during the first hours was very high — much higher, in facK, 

 than amongst the stai-ved flies for a corresponding length of 

 time. Again, the percentage of deaths in those experiments 

 where the poison-bait was left for more than one day was much 

 lower than was found amongst the starved flies for a period of 

 three days or more. It appears that when flies died from the 

 effects of feeding upon those baits in which the poison was more 

 concentrated they quickly absorbed sufficient to kill them, but 

 that this toolt a longer time with weaker poisons, unless the 

 bait was very attractive and they fed so greedily upon it as to 

 imbibe a sufficient amount of the poison to kill them in a shorter 

 time. This is apparently what happened with a 10 "/o formalin 

 bait, which was very attractive, and to which very many of the 

 flies came to feed at once and continued feeding for some time. 

 On the other hand, if they fed less continuously a longer time 

 elapsed before death took place. Probably the large percentage 

 of deaths amongst the survivors with clean baits, as compared 

 with the death-rate in the controls, was due to the poison pre- 

 viously absorbed in the first case. 



/Starved Flies. — From experiments on starving flies (Tables II. 

 and V.) it seems that they show more power of resistance 

 late in the year than in the autumn. For in the experiments 

 made on starving flies in November and December they remained 

 alive without food for 7-10 days, or even longer ; while in those 

 made in September and October all the flies were dead by the 

 sixtli day. The greater resistance shown by the winter flies 

 than by the autumn ones is what would be expected if they 

 have to pass the winter as imagines. 



* This is not counted, although it had the highest death-rate, hecause no recoi-d 

 was taken on the first day, and it is very probable that some of the flies died on 

 that day. 



