76 



cases cavities or canals were found in the larger spines, and 

 MohPs supposition that they are cellular formations which se- 

 crete oil is confirmed. These spinous formations are most re- 

 markable in the group Cichoracece ; they stand in single regu- 

 lar rows, and can only be observed with advantage from the 

 side. Among these spines of Cichoracece a connexion exists 

 which seems to be effected by a membranous mass clothing 

 the membrane ; the series of spines appear from this cause 

 like broad bands, and present, when observed from the side, 

 an umbraculiform appearance. 



In another class of vegetables, regular pentagonal and hex- 

 agonal figures, circumscribing walls, have their origin on the 

 outer tunic ; these are again interrupted as in Cobcea scandens, 

 and appear like piers of a bridge. In Geranium and Pelar- 

 gonium verrucous bodies are situated, like statues, on the 

 arches connecting the buttresses. (The history of the occa- 

 sional formation even of some imperfectly developed pollen 

 grains of these plants shows that all the elegant formations on 

 the surface of the pollen granules arise from similar erect gra- 

 nules in juxta-position, as they so frequently occur in the Mal- 

 vaceae. Rep.) 



In some plants the surface of the outer tunic presented the 

 appearance of net-work, as in Polemonium coeruleum, Gilia 

 tricolor, Metrodorea nigra, &c. 



A separate chapter treats of the intermediate bodies which 

 occur between the inner and outer tunics. They are most 

 easily seen in Astrap&a, and have there in an isolated state the 

 form of a plano-convex lens, of which the side turned towards 

 the outer coat is covered with granules resembling the external 

 surface of the outer coat. In other plants these granules do 

 not exist on the outer surface of the intermediate body, or 

 there is merely a simple ring present, as in Ruellia formosa. 

 The intermediate bodies are situated in general at those spots 

 where apertures exist in the outer tunic which are closed by 

 them, and these intermediate bodies have a far greater extent 

 than the apertures, on which account they cannot be consi- 

 dered as belonging to the outer membrane. In Alcea rosea, 

 where these apertures in the outer membrane are very nume- 

 rous, their inner surface is clothed with densely juxta-posed 

 globular granules, which are said to be the intermediate bo- 



