76 THE ENTOMOLOGIST 'S RECORD. 
particular part of the cave, not far distant from a place where the roof 
had collapsed, admitting a certain amount of light and offering an 
easy egress. 
The walls of the same cavern were hidden in places by dense 
crowds of a species of Blatta; the wingless females and larve of a 
Perisphaeria mined in the bats’ guano, the winged males mingling 
with the larger species on the wall. Thousands of small Stenopelma- 
tidae leapt about on the floor, and sat where they could find unoceu- 
pied room on the walls. All these appeared to feed on bats’ dung, and 
the Locustids were the prey of enormous Pedipalpi, which crawled in the 
open cave, groping for their food all round them with their feeler-legs 
crossed over their backs. Outside the caves, the smaller, more lightly 
built, cockroaches, were mostly diurnal, flitting about in the clearings 
and the sunnier reaches of the jungle. The common house cockroach 
of the country is Periplaneta australasiae. Large colonies of this 
species conceal themselves in the hollows of the bamboos of which a 
Malay house is principally built. But the most powerful forms, such 
as Panesthia, are hidden during the day in fallen trees and branches, 
into the rotten parts and crevices of which the different species of 
this genus are well adapted to insinuate themselves, by reason of the 
pushing power of their spiny hind legs, which are so strong that it is 
extremely difficult to hold the insects between the finger and thumb. 
On one occasion I found a specimen, probably P. javanica, in arounded 
chamber in the very centre of a great log, with several small white 
cockroaches crawling round it. The large individual seemed to be 
healthy, but only the jagged stumps of its wings remained ; it is 
probable that the others had eaten them off. ‘The colourless speci- 
mens were not all of the same size, and possibly they were immature 
forms of the same species as the other. Some large specimens in 
the Hope Collection at Oxford appear to have lost their wines quite 
as completely, only the same jagged edge remaining, but there is no 
information in their case as to how the mutilation came about. A 
number of forms, belonging to different genera of the Malay Blattidae, 
bear a more or less marked resemblance to wood-lice, and some may 
even be mistaken for the Crustaceans. As arule, such forms are 
found among dead leaves or under stones, in places which the wood-lice 
also frequent. Ido not believe, however, that this resemblance is 
mimetic, for it is hard to see how mimicry could benefit either party, 
or both, in this case; it seems rather to be adaptive ; wood-louse and 
cockroach, living under the same conditions, have developed the same 
general shape of body. The cockroaches, however, which had the most 
surprising habits of those which came under my notice, were certain 
aquatic forms belonging to the genus F’pilampra. While we were staying 
in a hill-clearing on the boundary of the States of Nawnchik and 
Jalor, the gelatine on some photographic plates, which were left to dry 
in a small hut built over the stream, was eaten away during the night. 
Our Malay servants assured me that the damage had been done by 
‘‘lipas ayer,” or water-cockroaches, but I did not believe them. A few 
days later one of them pointed out to me a cockroach crawling along 
the sandy bottom of a small mountain rivulet, and afterwards I secured 
specimens both from a jungle burn in Rhaman and from the Kelantan 
river. In the river the wingless females sit on floating logs, in the 
crevices of which they deposit their egg-capsules, just above the water- 
