100 Bulletin Wisconsin Natural History Society. [Vol. 8. No. 2. 



also notes of my own have enabled me to construct a flowering 

 curve for 122 of our species of Composites. Some time in April 

 or around the beginning of May, according to meteorological 

 conditions, the flowers of the earliest species, the dandelion make 

 their appearance, followed a few days later by Aniennaria neglec- 

 ta. During the first half of May two other species of Antennaria 

 begin to bloom, towards the end of the month two species of 

 Brigeron and one of Senecio arrive on the scene, and we notice 

 a gentle rise of the curve up to the beginning of July, from which 

 time on a very rapid rise leads to a maximum around the end of 

 August. From there on a more or less gradual decline is notice- 

 able, and the flowering season is, as a rule cut short by the appear- 

 ance of a severe frost around the end of October or the beginning 

 of November. According to the available data only 18 of the 122 

 species considered appear before the 1st. of July, the bulk of the 

 family is therefore made up of summer and fall bloomers. 



This curve agrees quite closely with the curve presented by 

 Robertson 7 for Carlinville, 111. 



Judging from the enormous number of individuals by which 

 these plants are represented in our flora we are led to consider this 

 type of floral structure a very successful one, one possessing a 

 rather favorable combination of characters. It is therefore not 

 surprising to learn that these flowers prove very attractive to the 

 flower-visiting insects, and that in some instances the relations 

 between flower and insect are quite close. Robertson* in referring 

 to the most important visitors of the Composite includes the Bom- 

 bylids and gives a flight curve for this family, according to which 

 the greatest number of species is on the wing from about the mid- 

 dle of June to the end of August when the maximum is reached. 

 A corresponding curve for the Bombyliida 1 of Milwaukee Co., 

 based on observations covering 22 species has its starting point 

 near the end of April, rises slowly to the beginning of July, and 

 from there takes a rapid spurt upwards to a maximum lasting 

 throughout the month of August. This curve bears much resem- 

 blance to the flowering curve of the Compositce. 



In the list of flowers for which the visits of Bombylids have 

 been recorded, and which are arranged according to families 

 (P- 95) 2 4 out °f tne 5 2 species or 46% belong to the Composite. 

 If we consider the number of flowers visited by each of the 18 

 species of Bombylids referred to in this paper, we have out of a 



7) Chas. Robertson. The philosophy of flower seasons etc. Am. Nat. 

 Vol. XXIX, pp. 97-117 (1895). 



8) Chas. Robertson. Loc. cit. p. 110. 



