Potassium Hydroxide hy Dicranura vinula. 401 
A large number of pupae, due to emerge in about a 
fortnight's time, were placed in a corresponding number 
of glass tubes. The head of each pupa was placed so 
as to be in gentle contact with the closed end of the tube, 
while the remainder of the pupa was loosely wrapt in 
cotton wool to prevent violent movement, and also to 
absorb any excrement that might trickle downwards. 
Behind the pupa was placed a firm plug of cotton wool, 
in order to give it a solid object against which to thrust 
the discarded pupa case. 
The tubes were arranged in rows on a board, which 
was slightly tilted, in order that the potassium hydroxide 
might flow down into the angle of the tube below the 
pupa head. 
By careful watching, as the time of emergence drew 
near, I was able to collect a measurable quantity of the 
fluid from the tubes, and by instantly removing the 
moths and placing a pipette to their mouths a further 
quantity was obtained. Many tubes had to be rejected 
in consequence of fouling by excrement, in order that 
the fluid analysed might be absolutely pure. As the 
liquid was obtained it was placed in a small tube, pre- 
viously carefully washed and dried, this was stoppered 
and kept in a corked bottle containing a little water, 
in order to prevent concentration by evaporation. 
These precautions were necessary, as the period of 
emergence spread out over more than four weeks. 
The amounts obtained were dealt with in two lots, one 
lot amounted to 0*6 c.c, the other to 0*4 c.c. I did not 
count how many moths had contributed, for I was 
never able to obtain all the secretion of any one 
individual, and the computation would therefore be 
worthless. 
The results of analysis (vide foot-note*) gave in the one 
instance r47 grammes, and in the other 1*40 grammes 
of potassium hydroxide in every 100 c.c. of the secreted 
liquid. 
* The method adopted was as follows : — The volume of secreted 
liquid was ascertained by scratching a mark on the tube at the 
surface of the liquid, and subsequently filling the tube up to the 
mark with water from a Geissler's burette, graduated to c.c, 
and provided with an Erdniann's float. The secretion was carefully 
washed out into a small beaker with distilled water, and the 
whole then faintly coloured with methyl-orange to serve as 
