238 Mr. E. Strand on new 



not only in the colour, as pointed out by previous authors, 

 but also in the shafjo, tiie network, &c. 



In the ancient collection of the Museum is one cocoon from 

 Caracas [Gollmer leg.) and one from Cuba {Gundlach leg.), 

 both, I am sorry to say, witliout the insects. The former is 

 15 mm. long and 8 mm. broad^ the supporting thread is only 

 7 mm. long, but has probably been torn off; the colour black. 

 The latter cocoon is 12 x 6 mm., the thread ca. 50 mm., the 

 colour white. In the Staudiuger Collection are three bright 

 salmon-red or orange-coloured cocoons from the Upper 

 Amazons — Fonteboa, S. Paulo, and Pebas, the first two with 

 the insects bred from them (Trickosiibas fonteboce sp. n., and 

 sancti-paulensis sp. n.) — and a white one from Merida. The 

 latter diflFers from the one from Cuba inasmuch as the 

 network is more fine-meshed ; the supporting thread is only 

 4 mm. long and at the end strongly enlarged in the form of 

 a plate, which is 2"5 mm. wide; I am not sure, however, if 

 this thread is entire. The species described below as Tricho- 

 stihas merida sp. n. belongs probably to this cocoon. The 

 sup])0rting thread of the orange-coloured cocoons bears fine, 

 perpendicularly offstanding fibrils, which are hardly to be 

 seen with the naked eye, and are as long as the diameter of 

 the thread. Owing to these fibrils tlie thread has a rough 

 appearance and easily clings to other objects. The thread of 

 the cocoon from Pebas bears, moreover, long, fine, woolly- 

 looking fibres, which are mostly parallelly directed. The 

 white cocoon from Merida and the black one from Caracas 

 have no such perpendicular fibrils at all; the orange ones, on 

 the contrary, bear such ones also on the heavier parallel 

 threads of the network, but none on the thinner cross-threads ; 

 woolly-looking fibres are never to be seen on the cocoons 

 themselves. The supporting thread is never fastened on the 

 inner side of the wall of the cocoon, but often distinctly on 

 the outer side. The cocoon from S. Paulo is especially 

 interesting, owing to the fact that the meshes are filled up 

 with a tissue of fine fibres, so that now but little of the 

 original network is to be seen ; this cocoon is also more 

 cylindriform than the others, 25 by 11 mm., the thread ca. 

 160 mm. long, while the thread of the cocoon from Fonteboa 

 is only 65 mm. The meshes form mostly an elongate 

 parallelogram, but those of the black cocoon from Caracas 

 are pentagonal or hexagonal ; it measures 15 by 9 mm. 

 The opening at the upper end of the cocoon is apparently 

 made, or at least widened, when the moth issues ; the net- 

 work around the opening is often denser than in the middle 

 of the cccoon. Blanchard says {I. c.) that the cocoons may 



