METHODS OF RESEARCH IN FLOWER POLLINATION 197 



To eliminate the influence of the Composition of the Flora upon the choice made 

 by insects, a standard must be found for the proportions in which the various flower 

 classes are represented in the flora of each month. The mere numbers of species 

 belonging to each class that are in flower would not be sufficient, for in this way 

 small, inconspicuous blossoms which attract but few insects would have the same 

 value given them as large, conspicuous forms, rich in nectar, and attracting many 

 insects. The actual standard indicating the importance of each class seems rather 

 to be the degree in which the seven flower classes are visited by all insects. The 

 number of visits that each class of flower received depends upon (a) the number of 

 flowers in bloom, (3) the number of individuals (frequency or rarity of species), 

 (f) the abundance of nectar, (</) the size and colouring of the conspicuous parts, 

 i.e. the totality of the means of attraction, the' physiognomic value of each class in the 

 flower world. 



MacLeod estimates for the species of each individual flower class the per- 

 centages of insect visits observed during a given month, and thus reaches approxi- 

 mately the degree to which the flower class in question attracts insects as a whole, 

 each group of insects being separately considered. If, for example (according to 

 Hermann Miiller's observations in the Alps), during the month of June, all the 

 flowers in the Alps below the limit of trees receive 947 diff"erent visits, these 

 according to MacLeod's calculation would be distributed as follows among the 

 flower classes: 



Total 947 99-8 



Of the 947 visits, 201 would be paid by allotropous Diptera (i.e. all the antho- 

 philous flies except Syrphidae, Bombyliidae, and Conopidae). If these were equally 

 attracted by the seven flower classes, the 201 visits would be divided between the 

 classes in the same proportions as the 947 visits of insects collectively were divided. 

 But as the following table shows the proportions are quite otherwise : the visits of 

 allotropous Diptera are paid as follows : 



Total 201 100-5 



