186 NATURE SKETCHES IN TEMPERATE AMERICA 



threads marked perfectly the course of her movements in my 

 absence. 



A Httle study of the various strands revealed the fact that 

 she had examined critically the tops of the buds and had dis- 

 covered in some way the failing nutrition of the flowers. She 

 had shown evidence of deliberation and had gone to and fro 

 in her travels over the bud tops, as was indicated by the various 

 threads. 



I have witnessed this spider change color occasionally from 

 white to yellow, during one of its moults. Packard^ has inti- 

 mated tliat this change of color is for the purpose of more 

 perfectly adaj)ting the spider to the flowers of the goldenrod, 

 and other yellow flowers that appear later in the season. From 

 what I have seen in nature, there is evidence that this s{)ider 

 has the power of changing color when visiting different flowers. 

 In the chapter on "The A.s.sassins in Lace," I have described 

 the behavior of this species in her white attire. Her habits 

 on this plant are somewhat different from those while she is 

 on the evening prinini-i . 



The Castle-buildixg Spider 



o 



\K autumn I explored a number 



of uncultivated city lots where 



weeds in ])rofusion had unbounded 



sway. - Patches of liigii grass, 



shepherd's purse, peppergrass, sedges, and ragweeds made the 



^ Journal of the New York Entomological Society. 

 - Observations taken in Chicago. 



