Note on the sexual generation and the development of the seed-coats 
in certain of the Papaveraceae 
CHARLES H. SHAW 
(WITH PLATE 15) 
During the past two years the writer has mounted and exam- 
ined material in some quantity of Sanguinaria, Chelidonium and 
Eschscholtzia, The facts in hand indicate that the embryo-sac 
phenomena do not widely depart from those described, for instance, 
in genera of the Ranunculaceae. The following are some of the 
facts observed. . 
Annual phases of Sanguinaria.—Material of Sanguinaria has 
been gathered at various seasons in order to obtain a complete 
view of the annual life-cycle. The flowers of the plant begin to 
be formed nearly a year before they appear, viz., in May of the 
preceding year, while the seeds of that season are becoming ma- 
ture. The stamens develop more rapidly than the carpels. By 
August the anthers are well formed, with four-lobed outline in 
cross-section, and the sporogenous tissue is beginning to be dif- 
ferentiated from that of the anther-walls. Cell-division proceeds 
during autumn, the size of the anther as a whole meanwhile in- 
creasing. The cells lying toward the periphery of the sporoge- 
nous mass begin to degenerate and by November the spore- 
mother-cells are seen as richly filled cells lying among the more 
or less disorganized tapetal cells. In this condition the winter is 
passed. Division of the mother-cells begins very early in the 
spring. Material killed on February 4, 1903, after several days 
of very mild weather, shows figures of dividing nuclei. Develop- 
ment of the spores and the final formation of the anther-wall seem, 
as might be expected, to be somewhat controlled by the weather, 
but are completed during this period, between the renewal of 
growth and blooming. The ovules begin to appear about August 
as undifferentiated outgrowths of the placental tract, but develop- 
ment lags behind that of the anthers. The integuments are not 
developed till the close of winter, at the time when the micro- 
429 
