ARES Fo PEE O dP PS ee Bae net oe ao : 
1877.] Botany. 563 
which, added to those already known, will aid in the study of this 
us.” i 
In Botanische Zeitung for October, 1857, Daniel Müller, of Upsal, 
gives an instructive account and a few figures of the incomplete flowers 
of certain violets. He states that the anthers contained only a few grains, 
which did not seem to him like perfect pollen, but rather like minute 
ovules, In spite of a whimsical theory with which he closes his paper, 
the account must be valued for the accurate descriptions of the incom- 
plete flowers of several species. 
A very interesting paper on this subject was read in July, 1860, before 
the Botanical Society of France, by M. Eugéne Michelet. The state- 
ments made by him agree essentially with those just given. 
In 1863, Von Mohl (in Botanische Zeitung) gives an abstract of the 
literature relating to cleistogamic flowers, and presents some instructive 
results of his own observations, of which the following is an abstract: 
“The process of fertilization in Viola elatior F. was more easily investi- 
gated than in the other species examined. In this species, as in all apet- 
alous violets, the style is short and hooked, and in immediate contact with 
the anthers with which it alternates. Besides these two stamens which 
in this plant I always found developed, I discovered in some flowers one 
ortwo more. Although the anther cells are only one seventh to one 
sixth of an inch long, they have a number of pollen grains. The greater 
part of these push forth pollen tubes even while they are still in the 
anther, and these tubes pass out of the upper end of the anther cells, in 
thick strings, directly to the contiguous stigma. If the stamens in a 
fresh flower are drawn away from the stigma, the tubes will not break, 
but the pollen grains will be released from the opened anther cell, so 
that the latter will be left empty. On tearing away the anthers, some 
pollen grains which have not pushed out any tubes will fall out from 
them. It appears to me questionable whether without such a mechanical 
Process pollination from these would take place; at least I have not 
observed any such case. With the drying of the anthers after fertiliza- 
Hon, the tubes in their course from the anthers to the stigma dry up 
also, and then break off when the anthers separate, without withdrawing 
the pollen grains which are there held fast. Similar appearances are 
Presented by the anthers in Viola canina, which touch the stigma. 
Besides this, it is seen that from those anthers, which in this species are 
always turned away from the stigma, pollen tubes start out and pass 
Wn ina serpentine course over the upper part of the ovary and the 
back and sides of the style. This observation is easily made by means 
of a Lieberkiihn illuminating mirror. In this species, also, I frequently 
pollen grains which had fallen out of the anthers, but I am not 
Sure whether this discharge takes place naturally; for if we examine 
anthers which have become dry after fertilization, and on which, there- 
fore, while fresh no force could have been exerted, they will be found 
