Miscellaneous, 1 55 



the cell-development than the manner of its fertilization. This is 

 especially true of that part devoted to the castor-oil plant, Ricinus 

 communis. In Ccdehogyne he insists that the true embryo does not 

 develop, but that the seeds proceed from adventitious buds from the 

 wall of the ovary. One may conceive of them as bulblets, analogous 

 to what we find in viviparous flowers ; but he does not explain how 

 cotyledonous seeds are constructed in this way. 



In 1894 the author of this paper decided to make for himself 

 careful observations on Ricinus. A number of plants were set out 

 in a mass where he could easily watch them. Every male bud that 

 could be observed was rubbed away before expansion. For all this 

 an abundance of seed matured. It was found subsequently that 

 there were other plants within a quarter of a mile. It was thought 

 best to repeat the observations another year. The following year 

 no plants were near but those under observation, and it began to 

 look as if the plants had no use for pollen. Still there seemed 

 doubt, and the observations were continued yearly to the end of 

 1897— all with the same result. Dr. Ida A. Keller, Professor of 

 Biology in the Philadelphia Girls' High School, an experienced 

 microscopist, aided me considerably in the observations made, and 

 only for the fact that in one of her acute examinations she dis- 

 covered a pollen-tube at work on the ovarium, there would have 

 seemed no room for doubt that pollen was unnecessary for the 

 production of perfect seed in Ricinus. It was determined to try 

 once more. The past year (1898) only one plant was allowed to 

 grow and only one flower-spike permitted to remain. All the others 

 were cut away in the bud. This permitted closer examination, 

 which was made almost daily. I noted for the first time that there 

 were on the pedicels with the female flowers buds scarcely larger 

 than pin-heads, and, under a strong lens, evidently antheriferous. 

 These were carefully cut away. As the female flowers opened they 

 all withered, unlike their fertility in former years. There seemed 

 no doubt that pollen was essential. After half the spike had fallen 

 away, the upper portion — say some twenty flowers — showed a 

 disposition to permanence. The pretty fringed pistils protruded and 

 continued perfect for many days. Finally, the capsules enlarged 

 until they had about reached the usual dimensions. Now it seemed 

 as if the plant would produce seed without the aid of pollen. After 

 turning brown and seemingly ripe they were found to be as empty 

 as Camerarius's mulberries were. 



"While under the impression during the earlier years of these 

 observations that Ricinus was fruiting in the absence of pollen, I 

 was encouraged by observations seemingly confirmatory by Judge 

 Day, of Buftalo, and Prof. Greene, of Washington, on other plants. 

 At the same time I had noticed that solitary female plants of GingTco 

 biloba were fruiting abundantly in Germantown, while the only 

 known male plant was at Woodlands, some ten miles away. 

 Eecently Japan botanists have discovered spiral coils of spermatozoids 

 in the generative cells of Gingko, and others have observed them in 

 Cycas revoluta. 



The conclusion reached by the author is that though for several 



