1922] ROBERTSON—FLOWERS AND INSECTS I51 
observers show bees as 53.2 per cent. This is a higher percentage 
than is shown in anthecological observations of any region, and 
indicates that observations were introduced which discriminated in 
favor of bees to the exclusion of other insects. My Flowers and 
insects shows for bee visits 43.4 per cent instead of 44.6. The 
difference comes from using some entomological notes which were 
separated from my anthecological lists and were never intended to 
be used as anthecological data. 
Dystropic visits.—Useless visits of insects should be con- 
sidered from the standpoint of the flowers and regarded as marks 
of imperfect adaptation. Lorw has called insects which make 
such visits “dystropic,”” and MULiER has called Bombus mastrucatus 
a “‘dysteleologue,” as if these insects were under some teleological 
obligation to make useful visits. So far as the flower is concerned; 
a bee collecting pollen from it without effecting pollination cannot 
count among the useful visitors. So far as the bee is concerned, 
however, the visit is legitimate and the flower must be counted 
among its pollen visits. In the case of a family of small bees 
(Halictidae) 165 visits, mostly for pollen, were observed to flowers 
which they failed to pollinate. Knuru, while omitting American 
visits under the plants to which they relate and redistributing 
them under the insects which make them, has indicated the so-called 
dystropic visits under conditions which make them entomologically 
irrelevant. 
EXAGGERATION OF FRAGMENTARY OBSERVATIONS.—Probably 
anyone who contemplates methods of publication will notice 
this as a common characteristic. Fragmentary observations, as 
regards publication, reviewing, or abstracting, get more considera- 
tion than they are entitled to, for the reason that it is easier and 
cheaper to do so. When it was complained that KnurH sup- 
pressed anthecological lists, the statement did not apply to the most 
fragmentary, the most worthless ones. His work gives 55 of my 
lists averaging 2.8 visits, and excludes 207 lists averaging 28.9 
visits. There is only one reason: the short lists are easier to copy 
and to print. 
One list containing 18 Syrphidae and 128 other insects is. 
omitted, while another list for the same plant, consisting exclusively 
of 5 Syrphidae, is given. Special mention is made of the occurrence 
