I50 BOTANICAL GAZETTE : [FEBRUARY 
difference is quite definitely shown in the first 100 cases in which 
collectors’ notes and anthecological lists are given for the same 
flowers (table I, 1). The latter show for visits: bees 44.1 per cent 
and flies 29.1, while the former show bees 86.5 per cent and flies 2.8. 
One indicates a rather miscellaneous assemblage, while the other 
shows such a strong preponderance of bees as to reveal the fact 
that most of the entomologists were collecting bees particularly. 
In the first 62 cases the bee visits stand 82.5 per cent, flies 4.2, 
against bees 33.3, flies 41.8, the maximum changing from bees to 
flies. In the next 38 cases the bee visits are 90.5 and 60.6, a 
difference of 29.9. These are Papilionaceae and rather melittophi- 
lous, so that the bee collectors’ notes give a more correct approxima- 
tion. In many cases they did not say whether the bees were 
collecting pollen or not, and often did not mention the sexes at all. 
In the case of 65 species of Compositae for which there were 
both collectors’ notes and anthecological lists (table I, 2) the 
percentages of visits stand: for the former, bees 79.2, flies 10.2; 
for the latter, bees 35.6, flies 37.0, the maximum again shifting 
from bees to flies. Comparing 59 species observed by MULLER 
(Fertilisation of flowers), 65 observed by anthecologists in Europe, 
and 85 observed by me in Illinois, there is more resemblance between 
the European and Illinois lists than between collectors’ notes and 
anthecological lists based on the same 65 species observed in Europe. 
In the case of 51 plants observed both in Europe and Illinois 
(table I, 3) the percentages of bee visits show: for the collectors 
88.6, for the anthecologists, Europe 42.5, Illinois 40.4. Here again 
there is more resemblance between anthecological observations for 
Europe and Illinois than between anthecological lists and collectors’ 
notes for Europe. As entomological data the collectors’ notes are 
just as unsatisfactory. In the three cases the percentages of visits 
of flies range: for the collectors, 1.4 to 10.2, with an average of 4.5; 
for the anthecologists, 9.5 to 41.8, with an average of 29.5. 
The insect visits to flowers recorded in my Flowers and insects 
are not included in KNutu’s work, but are broken up in a final list 
and redistributed under the insects, making a quasi-entomological 
subject of them, but they are vitiated by being mixed with collec- 
tors’ notes. These visits recorded in KNUTH (table I, 4) show for 
bees 44.6 per cent of the total. The visits recorded for other 
