462 
J. Osawa 
These irregularities in the pollen-development are not rarely met 
with in the parthenogenetic plants, as we know from the stndies of many 
authors. Juel (1900) found in Antennaria alpina that there exists scarcely 
any good pollen-grain at all in the loculus of the anther. Murbeck (1901), 
who has studied Älchemilla stated, that in A. alpina and A. sericata the 
disintegration occurs in pollen-mother-cells, even before the tetrad- 
division, while in A. speciosa and A. alpeslris, though pollen-grains are 
prodnced by the normal tetrad-di vision, their great part undergoes the 
degeneration, and that others, whieh look normally, have no germinating 
power at all. Rosexberg (1907) mentioned about Hieracium excellens, 
that several irregularities take place in the tetrad-di vision, though the 
heterotypic spindle is fornied. Winkler (1906) showed in Wickstroemia 
indica, that though the pollen-mother-cell generally undergoes the normal 
tetrad-division, the daughter-c-ells degenerate with out reaching the 
maturity, that many nuclei may be sometimes found in one tetrad as in 
T. albidum and also that the pollen-mother-cell often becomes directly 
a pollen-like body without any preliminary division. In Houttuynia 
cordata, studied by Shibata and Miyake (1908), there occurs no reducing 
division, but one homotypic mitosis, tlius producing onlv two cells, as 
often seen in T. albidum ; several irregularities, e. g., amitotic division 
of the daughter-nuclei of the pollen-mother-cell, the Scattering out of 
chromosomes in cytoplasm, etc. were also observed. 
Similar irregularities of pollen-development and the sterility of 
pollen-grains also occur very commonly in hybrid plants. For example, 
in the hybrids of Syringa (Juel, 1900), Cotton (Caxxon, 1903), Drosera 
(Rosexberg, 1904, 1909), Street peas (Gregory, 1905), Ribes, Bryonia, 
Syringa and Banana (Tischler, 1906, 1908, 1910), Oenothera (Gates, 
1907), irregularities and abnormalities were found, such as the early 
disintegration of the pollen-mother-cell, the Scattering of chromosomes 
out of the spindle into cytoplasm in the heterotypic division, amitotic 
division of daughter-nuclei, the formation of many nuclei in one 
tetrad, etc. 
In my present investigation it appears to be probable that tliose 
races, whieh eontain the larger number of chromosomes, were phylogeneti- 
cally derived from those others, whieh eontain less numerous chromo- 
somes. In T. platycarpum the tetrad division takes place quite normally 
in every respect; in T. officinale, according to Juel, there occur some 
irregularities, though the heterotypic spindle is normally formed, while 
in T. albidum many irregularities take place in the pollen-development, 
as mentioned before, and it is hardly possible to take such form as T. 
