AQUATIC COCKROACHES. 225 



if the supply is entirely cut off there is practically no reserve 

 supply contained in the body to draw upon. Hence the rapid 

 death of the insect when totally submerged ; if only partially 

 submerged death supervenes less rapidly, probably because some 

 air can be drawn through the thoracic spiracles. Panesthia 

 javanica is able to endure total immersion longer than the partial 

 immersion to which individuals were submitted, because these 

 individuals, when bound, struggled so violently as to make heavy 

 demands on their reserve air-supply ; their position was so con- 

 strained, so unusual, and so unnatural that they were not able 

 to "take matters quietly," even when fastened in a position 

 presumably favourable to drawing in a fresh supply. Plateau 

 has shown that aquatic insects drown more quickly than terres- 

 trial insects, and suggests that this is because their supply 

 of oxygen is quickly converted into C0 2 through their violent 

 struggles to escape, whereas terrestrial insects, when submerged 

 in water, soon cease to struggle, and, although they become 

 comatose, they recover power of movement when restored to 

 land. It would be of interest to learn if an aquatic insect such 

 as Dytiscus would endure partial immersion, i. e. with the tip of 

 the abdomen exposed, as well as the aquatic Blattids. 



Dr. Nelson Annandale discovered some aquatic Cockroaches 

 in the Malay Peninsula* ; the females were wingless, and rested 

 on floating logs, whence they dived into the water when dis- 

 turbed ; the males were winged, and were seen to rise from the 

 surface of the water, but were never seen to enter it. Dr. Annan- 

 dale states, moreover, that the egg-cases of this species were 

 found in crevices of the floating logs. If the Malay Peninsula 

 species belongs to the same subfamily of Blattidce as the Bornean 

 species, namely, to the Epilampr'ina, this discovery of egg-cases 

 is of some interest, for the Epilamprince are, so far as is known, 

 viviparous insects, the chitinous ootheca deposited by the females 

 of other subfamilies being represented by a delicate membrane 

 enveloping the eggs inside the brood-sac of the mother. Unfor- 

 tunately, Dr. Annandale's specimens appear to be lost, so that 

 they cannot be identified with certainty. 



Another species has been discovered in Formosa,! and Dr. 



* Ent. Eec. 1900, p. 76. 



f Shiraki, Ann. Zoolog. Japon. vi. 1906, p. 32, pi. 2, f. 4. 



