78 



sihly to changed conditions, initiates, carries on, and com- 

 pletes a change in which the insects are only valuable 

 assistants. 



As a means of solving the question, we have taken note 

 of the principal visitors of a hundred different kinds of 

 flowers, more or less. 



We find that it is a difiicult question to determine posi- 

 tively, owing to different conditions in different localities ; 

 the species vary, the seasons vary, and the insects vary ; 

 competition varies, and the daily weather varies ; but we 

 are able to say what flowers certain insects hav'e visited and 

 in which they appear to delight. Our attention has been 

 directed principally to the larger kinds : bumble bees, 

 honey bees, wasps, hornets, butterflies, etc., while not 

 entirely neglecting the minor insects. We will anticipate 

 our conclusions by saying that taking a general survey of 

 the subject we cannot And any marked preference of any 

 kind for any one color. 



In the Fall of one season, when golden rod and Joe-Pye 

 weed (Enpatorium) were at their height, these flowers 

 bloomed a))undantly along the side of a road leading through 

 swampy woods. Bumble bees were present in great num- 

 bers, visiting both of these flowers, one kind yellow, the 

 other i)urple. An actual enumeration was made, and the 

 number of bees on one kind was substantially the same as 

 on the other. The following year, in a different locality, 

 there were bumble bees found on Joe-Pye, but scarcely one 

 on goldenrod. The only way I could account for the differ- 

 ence was by supposing that the bees of the woods, l)eing 

 remote from flelds, were restricted to woodland flowers, 

 while in open flelds they would have a greater choice of 

 kinds. 



Symphoricarjjos vulgaris was discovered in Eastern Mas- 

 sachusetts. The Floras do not give the plant as growing 

 in New England, so that likely few, if any others, could be 

 found within two hundred miles. The flowers were monop- 



