BOTANICAL GAZETTE. 71 



originally proterandrous, became, from want of insects caused by its 

 early appearance, autogamous by synanthesis, and is now again, by 

 the adaptation to new insects, becoming dichogamous, but proterogy- 

 nous, instead of proterandrous.— Aug. F . Foerste. 



Mr. Foerste was kind enough to send with the above note dried 

 specimens of Erigenia, from which it appears that self-fertilization is 

 quite as.probable as in Hydrocotyle, but although there is no ques- 

 tion about the early elongation of the styles I could not be sure 

 that the stigmas were receptive until dehiscence was nearly com- 

 plete. From theoretical considerations, one might doubt the pro- 

 togyny. The development of synacmy in a dichogamous specie-; 

 might readily follow a decrease in the number of visiting insects, but 

 if their number afterward increased and dichogamy were again acquir- 

 ed, the hereditary direction of variation should tend to reproduce the 

 same form of dichogamy that had originally been lost, and that still 

 existed in related species and genera, especially in so uniform an order 

 as the JJmbellifercB. Mr. Foerste writes that he has noticed a similar de- 

 velopment of the stigmas in species of Thaspiiuu. It is to be hoped 

 that those who can obtain Erigenia will confirm this observation 

 another season, while Thaspiiuu, can be examined this year. Scandii 

 Pecten- Veneris should have been mentioned with Hydrocotyle in my 

 note in the Gazette for March, as being self-fertile, possibly through 

 incomplete dichogamy, its self-fertility being, as suggested by Mr. 

 Henslow (Trans. Linn. Soc.,2 ser. Bot., I, 365), correlated with the 

 inconspictiousness of its flowers. Apium is also stated by Mr. Dar- 

 win (Cross and Self-Fert, 172) to be self-fertile.— Wm. Trelease. 



Notes from the Jour. Royal Micr. Soc, February— 

 L. Guignard finds that the embryo-sac of the Mimoseae 

 (Acacia retinoides is taken as the type) is formed in 

 the following manner: At the summit of the nucellus, beneath the" 

 epidermis, an axial cell, larger than the adjoining one, divides into 

 two superposed cells, the apical and subapical. The subapical cell 

 enlarges rapidly and becomes segmented horizontally into three 

 superposed cells each equal in size to the mother-cell. Of these, the 

 lowest, the true mother-cell of the embryo-sac, enlarges at the ex- 

 pense of the others and the lateral tissue of the nucellus. Resorp- 

 tion soon commences in the two superposed cells; the nuclei lose 

 their sharp outline and soon disappear, together with the cell-walls: 

 finally the whole substance of these cells is absorbed in the develop- 

 ment of the lower mother-cell. This process is subject to variations, 

 but it is always the lower cell which becomes the mother-cell and 

 absorbs the others. 



The same worker finds polyembryony not an uncommon phe- 

 nomenon in the Mimoseae, especially in Schranckia uncinata, in 

 which it is allied with other abnormalities of structure. 



E. Wartmank finds that Spanish chestnuts germinate after an 



