666 SUMMARY OF CURRENT RESEARCHES RELATING TO 



The epidermis of the nucellus divides to form an epidermal cap 

 above the tapetum. Below the embryo-sac and during its development 

 the nucellus shows several threads of cells extending towards the 

 chalaza, which are crushed and in course of absorption. After fertilisa- 

 tion the nucellus is very slowly absorbed ; in the ripe seed of Pireae it 

 is still represented by a narrow layer of crushed cells, and sometimes 

 by a recognisable epidermis. 



Development in the embryo-sac immediately following fertilisation 

 is very slow, especially in the Amygdalese, but when the albumen is 

 formed it is, on the contrary, very rapid. The embryo always has a sus- 

 pensor, which is massive and short in the Amygdaleae, larger and formed 

 of a double or simple thread in all the other tribes. The lower cells of 

 the suspensor generally make a notch in the radicular extremity of the 

 embryo. In favourable cases (Pirus, Sanguisorba, &c), where the limit 

 between the cells of suspensor and embryo is well defined, the suspensor 

 seems to play no part in the formation of the organs of the embryo. 

 The early development of endosperm is uniform throughout the order, 

 commencing with a layer of free nuclei round the sac, which is then 

 transformed into a cell-layer by formation of cell-walls. Later the 

 cavity of the sac becomes filled, the Amygdaleae differing from the rest 

 of the order in that only the upper part of the sac is filled. The 

 nucellus is crushed and generally completely absorbed by the albumen, 

 which is' in turn in great part absorbed by the embryo. A portion of 

 the albumen always persists, though variously distributed around the 

 seed. The outermost layer is always remarkable by its abundant pro- 

 teid contents, but only in Rhodotypus, where the greatest quantity of 

 albumen is found, does it show marked differential characters from the 

 rest of the albumen. Owing to the generally small development of 

 albumen, and its intimate association with the seed-coat, the author 

 considers its function to be protective rather than nutritive. In con- 

 trast with the passive and mechanical protection afforded by the seed- 

 coat, it supplies an active physiological protection regulating osmotic 

 exchanges and preventing the intrusion of germs. 



The author considers that the ovule and seed show variations which 

 support the subdivision of the rosacea^ into a number of distinct 

 families. 



Existence of a Hypostase in the Ovule and Seed of Rosaceae.* — 

 Ph. Van Tieghem refers to his discovery of a small structure in the 

 pistil of seed-plants below the female prothallus, the object of which is 

 to arrest towards the base the longitudinal growth of the prothallus, 

 and which he calls the hypostase. It consists generally of a cupule of 

 isodiametric cells which have strongly lignified but not much thickened 

 membranes. When the pistil is inovulate, whether a nucellus is present 

 or not, it forms in the parenchyma of the carpel a single structure 

 common to all the female prothalli. When the pistil is ovulate, whether 

 the nucellus is permanent or transitory, it forms in each ovule a small 

 individual hypostase. Owing to its strong ligniuoation it resists all the 

 various diastatic agencies which are at work during formation of embryo 

 and albumen ; aud for the same reason is incapable of growth. Hence 



* Ann. Sci. Nat., xvi. (1902) pp. 159-60. 



