AND THEIR GERMINATION. 405 



radicle lies in a canal directed obliquely towards the base of the embryo (fig. 9), and is so 

 unattached that it may easily be taken out entire. But in Z. nana the radicle is short, 

 less oblique in its direction, and the orifice of the canal is not closed, as in Z. marina, 

 being covered only by a delicate semitransparent membrane (fig. 22). In Triticum, 

 Avena, and Sordemn, there is a tubercle immediately under the radicle, which looks not 

 unlike a closed orifice. 



As thus described, therefore, the embryo of Zoslera marina differs from those of Oryza 

 saliva, Zea Mays, &c., only in the cotyledon and its cauliculus l)eing curved by becoming 

 bent downwards, and in the smaller lobe of the radicle being removed from the base of 

 the cauliculus (fig. 8 a), which may be the effect of the pressure of the bent cauliculus, as 

 it makes a depression in the substance of the radicle. 



EriocaulonecB. — The germination of Eriocaulon septangulare is very singular, differing 

 from that of any other plant I have seen. The embryo first protrudes a horizontal 

 process, having a small speck on its outer part, consisting of a circular portion of the 

 testa, in the same way apparently as Mr. Wilson has described to occur in the germi- 

 nation of Lenina gibba (Bot. Misc. i. 145. t. 42) ; but in its after stages it differs in sending 

 up a leaf, which is directly away from the cotyledon, as in Juncacece and Liliacece, and it 

 cannot, therefore, be compared in this character to either Gramitiacece or Cyperacece. 

 Subsequently to the appearance of the first leaf, the horizontal process protrudes a root 

 from its under surface, which has no coleorhiza, although it most probably escapes through 

 a fissure in the tissue, as the subsequent radicles do. As growth proceeds, the horizontal 

 process becomes fissured by the enlargement of the root, which soon breaks through it, 

 and its remains are finally left as a root-like process on the root itself (Figs. 11, 12, 13, 

 14, & 15.) 



On referring to Klotzsch's figures of the germination of Pistia texensis *, I was agree- 

 ably surprised to find a satisfactory explanation of the germination of Eriocaulon ; and 

 no doubt it will also serve to explain that of Lemna, as figm-ed by Mr. WUson {loc. cit.), 

 which is admitted to be very obscure. In Klotzsch's figui-es the horizontal process is no 

 doubt the primary radicle, as it has so entirely the appearance of the primary radicle of 

 AracetB and their allies, and the same therefore must be the conclusion with regard to 

 the horizontal process of Eriocaulon septangulare (fig. 11) and the analogous process of 

 Lemna gibba. But Avhether this kind of germination is endorhizal or exorhizal remains 

 a question, because the growth of the radicle is arrested about the time the coleorhiza 

 first appears ; and possil3ly such a germination may be common to both endorhizal and 

 exorhizal plants, so that Fistiacea and Lenmacece may be, Kke Aracece, exorhizal. I 

 am obliged, however, to differ from Klotzsch in his description of the plumule of Pistla 

 texensis : what he calls the first leaf, I should undoubtedly regard as a stipulary process, 

 or perhaps only as the lips of the fissm-e in the cotyledon ; and what he calls the 

 second leaf must, of course, be the first leaf, which is figured as being directly away from 

 the cotyledon, as in Arum. 



CommelynacecB.—The endorhizal character is here well marked ; the coleorhiza is, how- 

 ever, very thin, and soon disappears. The first leaf of the plumule is directly away from 



* Uebcr Pistia (.\bhandl. der K. Akademic zu Berlin, aus dem J. 1852). Berlin, 1853. Taf. ii. figs. Q.R, S. 



