362 Animal Life 
damp hollow they do not pupate there, but walk long distances to find a suitable 
place in which to spin their cocoon. Some of mine, which I kept in a large bell-glass, 
were walking round and round continuously for several days, until at last they adapted 
themselves to circumstances, and made their open network cocoons among the old 
stems of the bedstraw. The pupa is beautifully marked in the same manner as the 
larva (Fig. 3). All my larve had gone to pupa by August 7th, and remain waiting the 
return of July. 
The first emerged July 3rd, almost before I was ready for it, though I had been 
watching it for some days; but it was too far advanced to attempt photographing it. 
My second pupa emerged during a short absence, and again I was disappointed; and 
though I watched the third for three days and had placed it in position for photo- 
eraphing, it seemed to take advantage of my absence for not more than three 
minutes—I had a look at it before I left my den—and on my return at 10 am. I 
found the pupa empty and the elephant crawling up the curtam! ‘To coax it on to 
my finger and then to its perch in a pot of peat (which had contained the pupa) 
and fix it in position did not take half a minute, and I had my first photo (Fig. 4), 
though mot so early a stage as I intended. It quickly crawled up, and after a little 
restlessness placed itself in a favourable position for Fig. 5, in which the wings are 
beginning to curl, as is usual at this stage, ten minutes after emerging from the pupa. 
Figs. 6 and 7 were taken at three minutes’ interval, the under wings being almost 
fully developed; they always appear to take precedence. 
At this time the elephant delighted in a number of wriggles and twistings, with the 
evident intention of exercising the muscles of its wings as it hung firmly Iss its claws. 
At 10.40 I took Fig. 8, when the wings were fully developed and only required 
to be flattened down. Hight minutes after this final movement took place (Mig. 9), 
