E. E. Atkin and A. Bacot 
523 
No. 6, this tube developed a vigorous bacterial growth; the larvae m£ide rapid 
progress, but aU died in the second or third skin. 
Distilled water. Of the eight tubes to which eggs were added ordy one showed 
infection by the tests used; on the 20th day of the experiment the position was as 
follows: 
No. 1, 2 eggs hatched, 1 larva dead, 1 living, both in first skin. 
No. 2, none hatched. 
No. 3, 3 hatched, 1 larva dead, 2 living, aU in first skin. 
No. 4, 6 hatched, aU dead in first skin. 
No. 5, none hatched. 
No. 6, 5 hatched, 4 dead, 1 living, all in first skin. 
No. 7, many hatched, probably aU, larvae aU dead but one; none past first skin, 
infected by mould. 
No. 8, 1 hatched; larva Living in its first skin. 
All the living larvae died within a week or two; the tubes containing dormant 
eggs were used for the under-noted experiment. 
On the 28th day of experiment a loopful of Living yeast cells {S. cerevisiae) was 
added to tube No. 1. Eggs which had been lying dormant for 28 days hatched 
within 5 minutes; rapid growth ensued and from time to time a fresh loopful of the 
yeast cells was added. 5 adults were reared and one specimen died in pupal stage. 
The time taken was from 30 to 50 days; the quantity of food given at five-day 
or weekly intervals—was a small meiss of cells scraped from a wort agar slope with 
a platinum loop—only allowing of development by fits and starts. 
Tubes Nos. 2, 3, 4, 5, 6 and 8 were infected by adding -5 c.c. of a yeast culture; 
larvae hatched from dormant eggs in most of the tubes within 5 minutes; others 
followed rapidly. 
All the larvae made rapid progress except those in tube No. 5; here moulds 
started to grow rapidly. Within a week there was only one larva living, and this 
one also died within a day or two. 
A small quantity, usually about -2 or -3 c.c. of the yeast culture, was added at 
intervals of several days or a week; the point it was wished to estabhsh being not 
the speed with which the larvae could develop on a diet of yeast cells, but the fact 
that they marked time during the intervals and spurted when the food was given. 
In spite of the rather crowded condition of the larvae in 6 or 7 c.c. of fluid, they 
remained healthy, except in tube No. 4 where moulds developed. In this tube all 
the larvae died. The result showed as anticipated the entire dependence of the 
larvae on the added yeast culture, their speed of progress being conditioned by the 
length of the periods between each addition to the tubes. In tube No. 2, 6 adults 
were reared, in tube No. 3, 6, and 6 individuals in tube No. 6; while in tube No. 8 
which contained more larvae to start with 9 adults were reared. 
Experiment XXIII. 
The increasing proportions of tubes in which sterilization broke down either 
from sporing species of bacteria or moulds, suggested that the contamination of 
the eggs laid in the breeding-pan was very gross. Steps were therefore taken to 
obtain eggs laid under more cleanly conditions and less hkely to be heavily infected 
with sporing bacteria and moulds (the method adopted is described on pp. 488-9). 
