1882.] 
419 
[Trelease. 
Hildebrand (1, 70-71) seems inclined to ascribe self-sterility 
to a set of guard-hairs at the margin of the stigma, which mechani- 
cally prevent pollen from touching the receptive surface. The 
first seems to me the more probable. 
Not a little has been written at different times on the struc- 
ture of Proteaceae as connected with their fertilization. 
Henschel (1, 78) mentions a number of species, of Mimetes, 
Adenanthus and Grevillea as exampfies of flowers in which a deep- 
lying fundus can be distinguished, beyond which the essential 
organs protrude, self-fertilization being hindered by the fact that 
the pistil is much longer than the stamens. Apparently neither 
the presence of pollen upon the stigmatic disk in newly opened 
flowers nor their dichogamy was noticed. Treviranus (1, 6) saw 
in the heaping of the pollen upon the stigmatic disk a sure means 
of self-fertilization, without recognizing the provision for cross- 
ing. Delpino (1, 181), with his usual accuracy, recognizes the true 
nature of the parts in the stigmatic apparatus, and divides the 
life of the flower into three stages, viz : that of dehiscence and 
the collection of pollen, that of the exposure of the pollen, and 
that of the maturation of the stigma. A lack of sufficient mate- 
rial prevented him from studying the last stage thoroughly, so 
he contents himself with expressing the probability that the nip- 
ple in the center of the stigmatic disk finally opens and either 
becomes papillose or deliquescent. An abstract of his conclusions 
is given by Hildebrand (2, 670). The storing of pollen and the 
protan dry are again mentioned by the same writer in a general 
consideration of the phenomena of pollination (2, 181; 162). 
Previously to this Hildebrand (1, 70-71) had described the true 
stigma as a papillose cavity in the center of the disk, its mouth 
being more or less completely closed by an especial set of hairs, 
the supposed use of which I have already noticed. More recently 
Mr. Bentham (1) has published the results of an exhaustive study 
of the floral structure in this order, from which it appears that 
several forms may be distinguished, corresponding undoubtedly 
to the mode of pollination. Unfortunately I have not been able 
to examine living specimens of these flowers, so I can do no more 
than refer to the paper. 
Several opinions have been published concerning the animals 
that aid in crossing the various Proteaceae, and the way in which 
