and on the Structure of the Seed in that genus. 185 



the hilum, we find the hollow space deep and narrow, the albu- 

 men appearing in the shape of a more compressed horse-shoe, 

 of which the crura are parallel ; and in the vacuity is seen the 

 same fungous substance in the form of a flattened plate, between 

 which and the integuments lining the cavity there exists a very 

 narrow space, showing that it is quite free from them in all 

 parts, except at the periphery of the flattened plate, where it is 

 attached to them by means of a very distinct cord or raphe filled 

 with numerous spiral vessels. This raphe is observed both in 

 the vertical and transverse sections, where it forms a broad 

 hollow tube partaking of the nearly circular shape of the cavity, 

 the inner surface of the tube being lined with a close web of 

 white spiral vessels ; it is seen to originate at one of its extremi- 

 ties in the hilar foramen of the outer tunic, and after making 

 nearly an entire circuit, it returns close to its starting-point, 

 where it disappears in the chalaza of the inner integument, over 

 the nidular place of the embryo. The embryo is very small, and 

 is best seen in the longitudinal section, where it is found im- 

 bedded in the albumen in the extremity of the lower leg of the 

 horse-shoe, or that nearest the hilar foramen of the outer tunic : 

 it lies somewhat obliquely, with its radicle pointing outwards to 

 a small external prominence of the albumen, and with the coty- 

 ledons directed towards the centre of the seed : the radicle is 

 terete, short, obtuse, and of the length and breadth of the coty- 

 ledons, which are pointed and much compressed, its whole length 

 being twice its own diameter, and about y 1 ^ th of the total length 

 of the seed. 



On examining this evidence, we readily discern the nature of 

 these several parts : the inference that the outer tunic, which 

 has hitherto been denominated the testa, is of extraneous origin, 

 and therefore of the nature of an arillus, is supported by four 

 reasons: — 1. Because it is not likely that the fungous plate, or 

 large funicular expansion found within the cavity of the seed, 

 could have found its way from the exterior into the hollow space 

 through the small hilar foramen of the outer coating; on the 

 contrary, it is more reasonable to conclude that this outer tunic 

 must have been a subsequent formation after the ovule had 

 become convoluted round that excentrically expanded support. 

 2. Because, had this tunic been the testa, that is to say, a simple 

 development of the primine of the ovule, it would have exhibited 

 some trace of its organic connexion with the other integuments, 

 especially about the region of the chalaza, and would have par- 

 taken of the same form, and have been drawn along with them into 

 the cavity. 3. Because this outer tunic is not only completely 

 free from all contact with the raphe, but is far remote from and 

 exterior to it, and therefore it must necessarily have been of a 



