ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 
407 
structure of the branches, in the difference of size or structure between 
the secondary branches or the leaves which spring from different sides 
of the primary branch. This dorsiventrality may be greatly modified, 
or even reversed, by changes in the illumination or in other external 
-conditions. Solar radiation may produce differences between the 
illuminated side and the side in shade ; gravity between the upper and 
the under side. The arrangement, number, and relative importance of 
the vascular bundles depend on the act : on of these factors. The dorsi- 
wentral structure appears to be the result of the coincidence of the 
illuminated side with the upper side on the one hand, and of the shaded 
side with the lower side on the other hand. 
Relationships of the Indefinite Inflorescences.* — Mr. E. A. N. Arber 
traces all the various forms of indefinite inflorescence back to the solitary 
axillary type of flower. From this have descended, in the first place, 
the spike and the capitulura, according as the primary axis has con- 
tinued to elongate, or has ceased to grow ; then, by growth of the 
secondary axes, from the former the corymb and the simple and com- 
pound raceme ; from the latter, the simple and compound umbel. 
Fruit of Gramineae.f — M. P. Guerin criticises unfavourably the con- 
clusion of Jumelle J that the caryopsis of grasses is simply an achene 
enclosing a seed without integument. H 3 states that, in general, the 
ovule has two integuments, composed only of two layers of cells. The 
•outer integument always disappears shortly after fecundation, while the 
inner one is persistent, and constitutes a true envelope to the ripe seed. 
In some cases ( Brachy podium , Bromus) the epid jrm of the nucellus takes 
part in the formation of the integument. In the pericarp the absorption 
is more or less complete ; the endocarp usually persists in the form of 
long isolated cells, or of a sclerified ring. With a few exceptions there 
is a complete fusion between the pericarp and the integument of the 
seed, justifying the retention of the term caryopsis. 
There are, however, some exceptional forms of fruit in the Gramineac. 
In Eleusine and Dactyloctenium the fruit is a true achene, with very thin 
walls. This is also the structure in Crypsis and Sporobolus, where the 
pericarp becomes, on maturity, transformed into mucilage. In Zizani- 
vpsis the pericarp is strongly sclerified, and encloses a free seed. 
Seedless Grapes.§ — Herr Miiller-Thurgau attributes the absence of 
seeds in some grapes to two causes: — (1) The pollen-grains are well- 
developed, but the ovules are not capable of impregnation ; either the 
pollen-tubes do not reach the ovules, or the ovule is itself sterile. To 
this class belong the sultanas and the currants of commerce. (2) The 
ovules are capable of impregnation, but the pollen-grains are degene- 
rated ; either the pollen-tubes do not germinate, or are incapable of 
impregnating the ovum-cell. Grapes which do not contain seeds are 
always smaller than those which do. 
* Journ. of Bot* xxxvii. (1899) pp. 160-7. 
t Bull. Soc. Bot. France, xlv. (1899) pp. 405-11; Journ. de Bot. (Morot), xii. 
(1898) pp. 365-74 (12 figs.) ; Ann. Sci. Nat. (Bot.), ix. 1899, pp. 1-59, 70 figs. 
t Cf. this Journal, 1889, p. 663. 
§ Landwirthsch. Jahro. d. Schweiz, 1898, 71 pp. and 4 pis. See Bot. Centralbl 
Ixxvii. (1899) p. 135. 
2 E 2 
