1903] NEEDHAM:— BUTTON-BUSH INSECTS 23 



tions, but other matters came into my hands demanding all my time, and no further 

 opportunity being now in prospect, I have concluded to publish such facts as are 

 already gathered. 



I studied the button-bush in a shallow open pond on the country place of Mr. 

 R. M. Bissell, within easy walking distance of my home, and in a "pot-hole" in the 

 woods near by. The pot-hole was filled almost exclusively with a dense and tall 

 growth of these shrubs. In the pond the clumps were scattered, low and broad in 

 form, and were restricted to the borders of a tlat island thnt lay in the midst of it. 

 This pond is of several acres extent, and is hardly too deep for wading with hip 

 boots anywhere. But it is being filled, by the land building of the button-bush 

 around the borders of the island, and by tliat of the tussock sedges that fringe all 

 its outer borders. Each summer it is reduced by evaporation to a few little pools, 

 at which time the button-bush clumps standupon a black soil that is fissured with 

 deep sun-cracks. On the island, clumps of glaucous willow and red dogwood crowd 

 the button-l)ush clumps in the rear, and they are generally fianked by some or all of 

 the following weaker plants: Cahunagrostis canadensis, Panicularia fluilans, Carex 

 ' utriiuJiia. Diilichinm spafJiace>im. Sparganium eurycarpiim, Sagittaria variabilis. 



The clumps here are impenetrable thickets. Muskrats buiid their huge hum- 

 mocks under them, immensely furthering the land-building process; tree frogs 

 climb the gray trunks, and red wing blackbirds build their nests in the tangled 

 branches. The density of their growth excludes other large plants, but an interest- 

 ing group of weaklings nestle in their shadows: Viola blatida, Scutellaria galericu- 

 lata, Galium trifidiim, and Onoclea sensibilis. At the surface of the water the 

 trunks are closely enwrapped by a moss of the genus Amblystegiian. 



The button-bush is our only woody representative of the madder family. 

 Coffee is akin to it, and most of its relatives are tropical shrubs. It is not well 

 adapted in some respects to our latitude. It develops only unprotected buds, that 

 are killed each winter together with the terminal shoots, the new growth from 

 adventitious buds in spring being late in appearing. One result of this, with a 

 bearing on the insect life found associated with it, is that the shoots of one season 

 are lateral to those of the preceding season, and the growth is scragged, and there 

 are many dead tops, broken ends, and exposed pith cavities. 



Very few insects have been recorded from the button-bush hitherto, aside from 

 the fiower visitors. Dr. Packard in the 5th report of the U. S. entomological com- 

 mission mentions three, two of these notes being citations of earlier records by 

 Harris and Riley. These are : — 



1. Platysamia cecropia (Linne), p. 402. 



2. Callosamia promethea (Drury), p. 525. 



3. Hyphantria cunea (Drury), p. 249. 



