10 PROCEEDINGS OF THE NATIONAL MUSEUM voi.. 71 



find how small a colony of plants would support colonies of these 

 caterpillars. The discovery of these very small isolated groups of 

 plants certainly speaks well for the pioneering capabilities of this 

 butterfly. 



In a moist hillside south of Otis Street, in Newton ville, there is a 

 small patch of turtlehead which certainly never consists of more 

 than 20 plants. While fairly close together, these are more or less 

 scattered in the long grass. For at least 29 years this little isolated 

 group of plants has supported a colony or two of these insects. 



About a hundred yards away down the hill there is another patch 

 of the plants, a little larger. Here also there are always to be found 

 a few colonies of this butterfly. The nearest colonies to these are a 

 mile or more away. 



We have often been surprised to find the webs on isolated plants 

 growing by the roadside far from any others. 



At the end of August, 1925, we made a search for webs in Essex, 

 as we wished to bring some caterpillars back with us to Washington. 

 But although we found the turtlehead in several places the few webs 

 we found were torn and ragged ones, containing only a very few small 

 and seemingly sickly caterpillars. 



The mystery was solved by the eventual discovery of several webs 

 quite different from the usual type of web found on the Chelone tops. 

 These webs were small, dense, and opaque, roughly fusiform, and 

 commonly about 4 inches long and an inch or so in thickness. They 

 somewhat distantly suggested a large, loose, and irregular cecropia 

 cocoon. Some were spun about several grass blades, one was on a 

 stem of Eupatoriuin purpureum, and others were on various plants. 

 They all agreed in being low down near the ground in the general 

 mass of herbage and therefore very inconspicuous. Sometimes they 

 were at the base of the Chelone stems, but often 2 or 3 feet or more 

 away from the Chelone plants. Frequently strands of silk ran 

 between these inhabited webs and the deserted webs on the summits 

 of the stalks of turtlehead. Probably there is always a silken trail 

 at first which, being delicate, soon gets destroyed. 



Presumably these were webs especially constructed for hibernation 

 after the caterpillars had finished feeding. 



In the locality in Prospect Hill Park, in Waltham, in the previous 

 year we had found the caterpillars prepared for hibernation in the 

 feeding webs, a part of which, usually the lower part, they had thick- 

 ened considerably. But this does not mean that there were not 

 plenty of small dense webs in the g rass or elsewhere which we over- 

 looked. 



One of the hibernatino; webs found at Es sex was just below a feeding 

 web to which it was broadly united by great numbers of silk threads. 



