Relationship between certain West African Insects. 449 
a stem bearing 22 Coccids. In a short time tiny black 
ants (identified as Pheidole rotundata, var.), which abounded 
in the house, found their way into the tube, which was 
then closed with very fine gauze and put away on a shelf, 
out of reach, 1t was thought, of more ants. However, 
more of the same species found it during the night and 
being unable to get in collected in a little knot on the 
gauze. 
“When the tube was inspected at 8 a.m. on the following 
day, 14 of the Coccids had disappeared—most of them 
entirely. The rings which had formed the basal portion 
of the scale of a few Coccids were, however, left by the 
caterpillars. 
“Later in the day I actually watched with a lens one of 
the larvae eating a Coccid, and at 1 p.m. only 3 Coccids 
out of the 22 remained. The larvae passed frass abun- 
dantly. The ants took no part in eating the Coccids.” 
A note dated Jan. 15, 1912, is as follows :— 
“ At 5 p.m., Jan. 14, 1912, the 2 larvae were placed in 
separate boxes and all ants excluded for 24 hours. By 
5 p.m. on the following day A had consumed 12 out of 
the 15 Coccids that I had placed at its disposal, and larva 
B had taken 16 out of 28, a few basal portions still remaining 
attached to the stems supplied to both larvae. I found 
that the larvae would eat these Coccids whatever the 
plant they happened to be attached to. The secretion 
of the Coccids was not sweet to the taste, but had an 
aromatic flavour rather suggestive of turpentine. 
“These larvae presented the same general characteristics 
ar those of A. vininga, being oblong on dorsal view with 
lateral surfaces sloping downwards and outwards. They 
had a hard tough toad-coloured skin covered with coarse 
rough tubercles, evidently protective in function, and it 
extended down as a fold on all sides in carapace fashion 
so as to protect the softer lateral and ventral surfaces. 
The lower margin of this fold bore a fringe of very fine 
hairs such as would efficiently prevent small insects from 
crawling in underneath. The segmentation characteristic 
of Lepidopterous larvae was shown only by the presence 
of spiracles, but rather more than halfway to the anal 
extremity was a deep transverse groove, the only region 
at which, owing to the leathery consistence of the cuticle, 
it was possible for flexion to take place. The cuticle was 
indeed so hard that a larva placed on its back was unable 
