.89,. OUR KNOWLEDGE OF SEEDLINGS. 755 



The walls of the seed are sometimes thickened at the chalaza, or 

 point where the vascular bundles enter the seed-coat, and this 

 may cause a notch in the apex of the seed-leaf ; similarly, auricles 

 at the base are explained by the effort of the cotyledon to fill up 

 the space round the radicle. A section of the seed of Ciiphea 

 silenoides illustrates both these phenomena. In the convolvulus and 

 its tropical allies, the ipomceas, a ridge at the chalaza is sometimes 

 so prominent that the cotyledons cannot grow in length, and are 

 consequently divided almost to the base. 



Sometimes, however, as in Bevbevis aqnifolinm, the cotyledons are 

 notched though lying quite free in the endosperm. 



Inequality of the two cotyledons may be due to folding in the 

 seed, as in the radish or cabbage (Fig. 18), where they are applied 

 face to face and then doubled longitudinally ; the outer one, having 

 more space, is the larger. In the geranium there is a want of 

 symmetry in each cotyledon, due to one-half of each being folded 

 inside one-half of the other, the two inner halves being the smaller 

 (Fig. 19). 



The curiously divided cotyledons of the lime are explained by 

 their manner of growth in the seed, where they meet the wall and 

 then curve back, following its general outline ; to fit the walls neatly 

 without crumpling they must be cut into lobes (Fig. 20). 



These are only a few of the considerations arising from the de- 

 scriptions and illustrations, which form the bulk of the work. 



It may be of interest to notice a few of the many other points of 

 interest revealed in a perusal of the systematic portion. 



In the Ranunculaceae the embryo is minute and embedded in 

 a copious endosperm, and we do not find the great variety in shape 

 which usually obtains where the cotyledons have to be packed in a 

 given space owing to the embryo occupying the whole seed. 



The prevalent type of cotyledon is broad and bluntly ovate or 

 oval; Ranunculus avvensis quoted above (Fig. g) is an example; 

 ■Ceratocephalus falcatus (Fig. 8), with its linear cotyledons, is a marked 

 exception. The genus Anemone shows the greatest amount of modifi- 

 cation ; the petioles of the cotyledons are more or less connate in 

 most species at the base, but in A . coronavia half-way up, in A . vupicola 

 nearly to the top, in A. polyantha quite to the top with the blades also 

 somewhat connate at the base ; m A. nemorosa the petioles are quite 

 suppressed. Delphinium affords an instance of an interesting relation 

 between length of hypocotyl and presence or absence of petioles in 

 the cotyledons. Thus in D. staphysagvia the cotyledons are almost 

 sessile, but the hypocotyl is three centimetres long, in D. datum, on 

 the contrary, the hypocotyl is very short, but the cotyledons have 

 long petioles ; in both cases the elevation of the cotyledon is assured. 



This relation is also noticed between two species of Vitis, and 

 again between Eucalyptus mavginata and E. calophylla. 



Clematis may be cited as an instance of the occurrence of aerial 



3B 2 



