474 RECORD OF CURRENT RESEARCHES RELATING TO 



Two primary jiro-embryo segments are formed ; but, notwithstanding 

 Hofmeistor's assertion, there cannot be said to be any considerable- 

 sized susjiensor ; but the division of the terminal cell takes place in 

 quite a different way from what it does in the Cruciferae, Eanun- 

 culaceae, and Labiataa. 



With the exception of the apparent absence of the "comites," 

 L. polypJiyllus agrees in essential points with L. varius ; L. mutabilis 

 has also many points of similarity, and the structui-e of the ovule 

 before fertilization is essentially the same. After impregnation, on 

 the contrary, a central string of protoplasm is found in the proto- 

 plasmic layer of the embryo-sac, which runs to about the middle of 

 the ovule, and then puts out four or five lateral strings. In this 

 system of strings are the nuclei of the secondary cells which are subse- 

 quently formed from them by transverse septa, and which may still be 

 recognized in the rijoe seed. The ovum lies at one of the points 

 where the system of strings branches. In Lupinus luteus the processes 

 differ yet more from the normal ones in Angiosperms. 



The formation of the endosperm in Lupinus also shows pecu- 

 liarities which apparently do not occur elsewhere. A distinction 

 must be drawn between the primary general endosperm, the nuclei of 

 which are distributed over the inner wall of the embryo-sac, and the 

 true " endosperm-body," which occupies only a part of the cavity near 

 the embryo. The formation of the numerous nuclei of the general 

 endosperm begins near the micropyle, possibly by division of the 

 primary nucleus of the embryo-sac and not by free cell-formation. 

 The further formation of the endosj)erm-body results from a small 

 portion developing into several layers accompanied by multiplication 

 of the nuclei. A parenchymatous tissue is then produced by the 

 formation of division-walls within the endosperm. These later 

 processes differ essentially from those observed by the author in the 

 Papaveracese, &c,, since in the latter a peripheral layer of cells is 

 first of all separated, which then breaks up by division into radial 

 rows of cells. 



Embryogeny of the Orchidaceae.* — The physiological processes 

 which go forward in the embryo have been made the subject of 

 investigation by M. Treub in the case of the Orchidaceae, which 

 present great facilities for this purpose, from the absence of endo- 

 sperm and the simple structure of the embryo. 



The absorption of nutriment takes place ordinarily through the 

 suspensor (pro-embryo), which is xmusually developed, and some- 

 times presents special adaptations for this purpose. The true 

 embryo is provided with a thick cuticle which prevents the absorption 

 of food-material by the surface. This was proved by experiments 

 with osmic acid. 



In many species, as Orchis latifolia, 0. pyramidalis, &c., the 

 suspensor consists of a simple row of cells, is of great length, pro- 

 trudes beyond the exostome, and becomes applied to the j)lacenta3 

 which abound in starch-grains, oil-drops, and other assimilable 



* 'Verh. k. Akad. Wetensch. Amsterdum,' xix.; see ' Bot. Zeit.,' xxxviii. 

 (1880) p. 57. 



