128 
Psyche 
[August 
jar was then introduced a small glass beaker containing a bit of 
carefully inspected fish meat. The jar was then closed by having 
mosquito netting tied over its open end and by setting in place 
on this netting the glass cover of the jar without, however, 
clamping it down. Thus it was believed that the jars were 
effectively protected against the introduction of eggs from the 
outside and that the gases generated by the decomposing meat 
within could escape. Fluid could be introduced into the jar by 
removing the glass top and pouring it in through the meshes 
of the netting without, however, allowing the accidental en- 
trance or escape of flies should there be any at hand. The fish- 
meat in the jars was not sterilized by cooking, for it was found 
to decompose much more freely and satisfactorily when un- 
cooked. 
The fifty jars thus prepared were set aside April 2, 1919, 
and allowed to stand twenty days. After this period a searching 
inspection showed that none of them contained maggots. These 
would surely have been seen had they been introduced by ac- 
cident with the meat. On April 23 twenty-five jars were infected 
each with a single fly egg and the remaining twenty-five were 
held unchanged as checks. From time to time during the next 
few weeks a small amount of distilled water was poured into 
each jar but otherwise the jars remained closed for this period. 
On May 14 the contents of the jars were examined. The twenty- 
five jars used as- checks contained no evidence of flies. In the 
twenty-five infected jars seven were without maggots or pupae 
and eighteen contained each a single pupa. Thus there was no 
evidence of increase. A second trial carried out in the same way 
in May and June yielded similar results. 
In the autumn of 1919 tests were resumed. These were of 
two kinds, one to ascertain what hatched from, an individual 
egg and the other to find out what came from the larva. To de- 
termine what came from a single egg, 'twenty eggs on October 29 
were put with a few drops of water each into a separate Syracuse 
watch-glass. Twelve of these had hatched by October 30 and 
each produced a single larva. Eight failed to hatch probably 
because of mechanical injury. Between October and the end of 
