RED-CLOVER SEED PRODUCTION. 3 
presence or absence of insects influence seed production. The work 
of Sirrine (37, p. 89-90), as well as that of Witte (46), showed red clover 
to be self-sterile. In the experiments of Cook (5), Shamel (36), and 
Kirchner (21) no seed was produced when heads were covered before 
blooming and not pollinated. Fruwirth (11; 12, p. 163-166) did not 
obtain a single seed when heads under cover were left undisturbed 
or when they were pollinated with pollen from another head on the 
same plant, while heads pollinated with pollen from another plant 
produced seed. Bolley (4) obtained but two seeds from one head 
of a large area which was placed under a fine screen before any of the 
flowers came into bloom. He states that insects other than bumble- 
bees must pollinate the flowers, since the bumblebees were scarce 
and the clover set well. Genevier (13) states that the fertilization 
of clover does not depend on the presence of bumblebees. Pammel 
and King (32) report but two seeds from 643 heads which were 
allowed to mature under a screen cover, while Washburn (45) says 
that only by the aid of bumblebees was he able to obtain seed. 
Armstrong (1), in writing about New Zealand, says there is every 
reason to believe that numerous individuals belonging to Trifolium 
pratense are self-fertile and that they produce self-fertile progeny. 
_ According to him the American strain is usually, if not always, self- 
fertile. McAlpine (24) discusses Garton’s experiments, which show 
that the self-fertilizing property is as common with red clover as it is 
with the bean. The following is quoted from Kerner (20, p. 407): 
‘“Pisum and Ervum, Lotus and Melilotus, the various species of Tri- 
folium, almost all of them, when unvisited by insects, ripen seed, 
only a few species here and there being infertile when dependent 
upon their own resources.”’ Nothing definite can be taken from 
Kerner’s statement, since he does not quote any species or give defi- 
nite exceptions to his statement. Hopkins (14) says he is not ready 
to admit that self-fertilization does not take place and that he is 
’ inclined to believe a crop of seed can be grown without the aid of 
bumblebees. The same author (15, p. 73) states that honeybees 
serve the same purpose as bumblebees in cross-fertilizing red clover. 
The work of Beal (2, p. 325-328) shows that bumblebees increased 
the seed production about four times, since in a check cage he received 
25 seeds from 50 heads, while in the cage where bumblebees were 
placed 94 seeds from 50 heads were obtained. Martinet (27) found 
red clover to be self-fertile, stating that cross-pollination might 
have been brought about by very small insects (undoubtedly mean- 
ing thrips). Fruwirth (12, p. 163-166), however, showed that 
thrips transferred from other clover fields in large numbers produced 
no seed in his experiments. Meehan (28) states that a careful exami- 
nation of the clover flower in all its stages convinced him that from 
its structure and behavior it was self-fertile. It is still an open 
