WATER-PRODUCING INSECTS 91 



flowers. Strong fibre used as thread is obtained from the 

 leaves, the name of the plant being indeed that used for 

 " thread." The tall flower-stalks of these aloes and agaves 

 form quite a noticeable feature in the Imerina landscape in the 

 early summer. In the orchards, soon after the mango has 

 finished flowering, we may see the curious whitish flowers of the 

 rose-apple, a sort of ball of long stamens, showing conspicuously 

 among the foliage. 



It is well known by those who live in Madagascar that there 

 are, at certain seasons of the year, a number of insects found 

 on trees which produce a constant dropping of water. Happen- 

 ing one day to be standing under a peach-tree in our garden 

 from which water was dropping, I found that there were 

 clusters of insects on some of the smaller branches. In each 

 cluster there were about twenty to thirty insects, and these were 

 partly covered with froth, from which the water came. The 

 insects producing this appeared at first sight to be small beetles, 

 about half-an-inch long, black in colour, with golden-yellow 

 markings on the head and thorax, while on the wing-cases there 

 was a chequer of minute spots of yellow on the black ground. 

 After observing a single insect for a few seconds, I noticed that 

 the tail was quite flexible and moved sideways, and was con- 

 stantly protruded and then withdrawn a little, and it was 

 evident that these little creatures were the larval form of a 

 species of beetle. The sap of the tree is extracted in such 

 quantities as to maintain their bodies in a state of saturated 

 humidity. The activity of the larvae seems to increase as the 

 heat of the day progresses, and to diminish again towards 

 evening. But the object of this abstraction of fluid from the 

 tree, and the purpose it serves, is still a subject needing investi- 

 gation. I have observed these insects on other trees mangoes, 

 acacia, zdhana, and others ; they appear indeed to be very 

 common, and the ground underneath the branches where they 

 cluster is covered with small patches soaked with water. A 

 French naturalist, M. Goudot, described an insect apparently 

 of the same kind as that found in Imerina as the larva of a 

 species of Cercopis, and nearly related to the cicada of Europe. 

 The quantity of water produced from a tree at Tamatave seems 

 to have been much greater than that observed in the interior, 

 and resembling a small rain-shower ; probably this was due to 



