6 
Bulletin of the EANHS 28( 1 ) 
path lit by the sun from directly overhead. The weather 
during this period included some good sunny days, as 
the cool wind finally died down around the middle of 
September; but the proper rains didn’t come until early 
November. Everything was up to a month or more later 
than it used to be (before the days of the greenhouse 
effect?!) 
FOOD PLANT 
But what was its food plant? Van Somcren's (1974) list 
of food plants gave information on two Cymolhoe species. 
The one (C. coranus) feeding on Kigelia didn't sound so 
plausible for a forest species But C caenis, on two 
species of Rawsonia trees sounded better. Williams 
1 1969) gives Dovyalis as one food plant of C. iheobene— 
a butterfly which looks superficially similar to C. teiia. 
And, as both Rawsonia and Doxyalis are members of 
the same family, Flacourtiaceae, that seemed a hopeful 
family to track down. 
According to the Utafiti write-up of the Taita hills 
forests survey (NMK. 1989). there are four species of 
Flacourtiaceae around the Taita Forests. Dovyalis 
(related to the Kai Apple), was rare at Ngangao. but not 
at Mbololo or Chawia. Trimeria was again only found 
near Ngangao (although I have seen it at Mwambirwa, 
and it Is relatively common, for example, at Mbale. 
which is not forest at all). That left two medium-sized 
trees, Aphloia and Dasylepis, both of these were recorded 
as ‘occasional’ in Ngangao and Mbololo. Both are also 
restricted to either Taita, or the other Eastern Arc 
Forests. These two looked possible food plants then, for 
the Taita Glider. 
Following a couple of visits to the East African 
Herbarium, I was hopefully armed with enough 
information, diagrams and visual memories to recognise 
either of these two genera when encountered. It didn’t 
lake long to find Dasylepis. It is fairly distinctive with 
its large green leaves, pointed at the end, on short 
petioles, borne alternately on woody stems The upper 
surface of the leaves is dark green, while on the 
underside, the raised veins loop across to each other 
near the outer smooth leaf margin. The racemes of 
reddish spherical flowers or fruits, with small petals, 
were also characteristic, as sometimes was the flaking 
reddish bark in older trees. And when the leaves are 
new and fresh, they are sometimes coloured a coppery 
red— especially under brighter light conditions. This 
contrasts with the older leaves, which in the forest itself 
are often heavily encrusted with a rich growth of algae, 
lichens and mosses. The Dasylepis trees in certain areas 
are certainly very ‘occasional’ . but in other places they 
are very common as an understorey layer, sometimes 
several trees almost directly adjacent to each other. On 
10 September, a long search revealed no larvae; but there 
were pupae— two empty and one still occupied. This 
was completely green, under the leaf, supported by the 
tail end so that the body was parallel with the leaf. On 
21 September, the adult butterfly emerged— it was a 
female Cymothoe teita. which was released back to 
Ngangao after a few days when I couldn’t find it a mate! 
At this point I felt the food plant question was more or 
less solved. 
LARVAE AND EGGS? 
During November, I found a total of three caterpillars 
filling van Someren's general description of 
Nymphalidae caterpillars as ‘cylindrical, with branched 
spines’ All three were found on the same (or 
neighbouring) Dasylepis trees, and were 19 or 20 mm 
long. They were basically green and cylindrical (even 
square in cross section)— but the head was orange/brown, 
slightly wider than the body, and there were two whitish 
longitudinal stripes, one on each side, running along the 
length of the body at the upper angle/edge of the 
‘ cylindrical ’ shape. The lower half of the body had a 
skirt of pale hairs, while most segments had a pair of 
branched hairs arising from near the white longitudinal 
stripe. Most hairs (as long as the body was wide) ended 
with a double prong, those on the first two and last one 
segments, were three-pronged (Neptune’s fork). At the 
front the caterpillar could effectively use these to deter 
anything disturbing it by swinging its head rapidly 180° 
from left to right. 
The caterpillars tended not to wander very much, 
often staying on die one leaf, eating it away from the 
end. until virtually left only with the petiole. One of 
these diree eventually reached 28 mm, and later pupated 
While it didn’t pupate perfectly— the pan of the larval 
skin didn't come off— it was similar enough in shape, 
size and colour, to be the Cymothoe. 
I found three batches of eggs during November. These 
were laid under the leaves, in groups of 19, 35 and 37 
respectively. The leaves were either old or fresh green 
and from 1-3 m above the ground. Each egg was around 
0 9 mm, white, beautifully spiky all over (like a 
radiolarian). They were laid in short rows, each separated 
from its neighbour by about half an egg’s width. From 
the first two, relatively few hatched (three and nine 
respectively); and these were all dead within three days. 
Part of the problem was drying up. despite my attempts 
otherwise So, the larger hatch (collected on a separate, 
later occasion) was kept on a larger bit of plant, and 
kept as moist as possible. This produced a roughly 90% 
hatch. The emerging caterpillars were about 2 mm long, 
pale, slightly hairy (enough to get them stuck in any 
condensation!) and with a wide dark head (which was 
visible through the shell about a day before they hatched). 
The very first batches of caterpillars had died, even 
when living on what seemed a young delicate piece of 
leaf— there were no obvious signs that they had even 
tried to eat it. When the third lot of eggs hatched, after 
a few hours altogether, they then dispersed completely 
and it was unusual to find any two together. This capacity 
to disperse completely (when conditions were 
unfavourable) was only matched by their ability to 
reconvene— in this case, on some very fresh shoots. They 
collected on the smallest, freshest leaf, which they then 
consumed And even then, when disturbed, they could 
show the head waggling which was hopefully a sign to 
indicate that they could prove to be C. teita. After five 
