A. REIJNE, A COCOONSPINNING THRIPS. 4 1 



out cacaothrips-larvae, both in the held and laborator)'. 

 Sometimes they took other small larvae (of Homoptera) and 

 once it was observed, that an adult cacaothrips was attacked. 

 Their importance as natural enemies of the cacaothrips 

 however is small, for they take no more than one larva per 

 day and their number compared with that of the cacaothrips 

 is very small, perhaps no more than o.i ^/q. 



The most peculiar habit, which to my knowledge has 

 never been described in the case of Thj'sanoptera, is the 

 spinning of a cocoon before pupation. This cocoon was 

 always made in the angle, formed by the midrib and lateral 

 veins of the leaf (fig. 4). Once it was spun in a curled-up 

 tip of a piece of blottingpaper, laid in the Petri-dish, in 

 which the thrips was bred, to keep the air moist. Of ten 

 cases, in which pupation was observed, once it happened 

 without making a cocoon and once a larva, that begun 

 spinning (fig. 4), broke its threads, when the light was too 

 much concentrated on it while under the microscope, and 

 pupated free on the blottingpaper. 



The spinning was begun by attaching some threads between 

 the veins (fig. 4). The larva, enclosed under this web, in 

 about 24 hours makes a dense cocoon of it, which is also 

 closed at the bottomside, as is seen when it is detached 

 from the leaf. After it is ready, the insect pupates. No 

 prepupa-stage (a stage with wingcases and free antennae) 

 was found as in other thrips. A cocoon with a pupa in it, 

 mounted in Canada-balsam, showed only one cast skin. This 

 was the skin of the larva, as was evident by the structure 

 of the antennae. It showed, that mandible and maxillae were 

 also cast in this molting. P. BUFFA (,,Redia" VII, p. 71 — 107) 

 has found for HeliotJirtps liaemorrJioidalis BOUCHÉ, that the 

 mouthspines are only once shed, namely at the end of the 

 larval period in changing to prepupa. This full}' explains, 

 that during the prepupa- und pupa-stage no food is taken 

 and the insects remain motionless when not disturbed. Many 

 of m)' observations confirmed this fact. The same was found 

 in HcliotJirips riibrocinctits, where during the whole larval 

 period strongl}' chitinized mouthspines with brown tips are 

 seen, even in embrs'os prepared out of the leaf. In molting 



