128 



I therefore communicate the results of my observations on 

 this subject, which are somewhat different from those of M. 

 Schleiden. The membranous cylinder from which M. Schlei- 

 den starts is not always present, but where it occurs cellular 

 nuclei are formed in it and cellular walls around them, and 

 thus the cylinder is divided into cells ; the terminal cell, how- 

 ever, expands in a globular form, and from this is 'developed 

 the entire embryo, which sooner or later separates from the 

 articulated thread originating from the older parts of the cy- 

 linder and termed by De Mirbel " bearer ;" it does not belong, 

 as it appears to me, to the essential parts of the embryo. Thus 

 the embryo appears on its first occurrence to be a simple glo- 

 bular cell, and has consequently the form of the most simple 

 of plants. This globular cell forms, from within outwards, into 

 a cellular mass, and it is only after it has attained a certain 

 size that the cotyledons are developed ; but at the same time 

 a growth of the globule takes place towards the opposite end, 

 and consequently the formation of a true axis. I confess, how- 

 ever, although I regard the stem as that part from which the 

 leaves proceed, and have anatomically demonstrated it, that I 

 could not observe the apex of the embryo-globule (punctum 

 veyetationis, Wolff.) to remain free between the projecting 

 cotyledons, I only saw the appearance of the first trace of the 

 future plumula at a subsequent period. 



M. Kunth has shown* by observations on the seeds of 

 Crucifera that the embryo of these plants adopts only towards 

 the ripening of the seed, in consequence of external causes, the 

 various forms which it exhibits in the perfectly developed 

 state ; and that those differences, since the causes which give 

 rise to them must always re-occur in the same plant, afford 

 very constant and important characters. He found in the 

 young seeds of Erysimum cheiranthoides, L., and E. officinale, 

 L., the radicula always situated laterally to the cotyledons, and 

 these only became incumbentes with advancing maturity. In 

 young seeds of Raphanus sativus the embryo was still nearly 

 straight, merely curved at the radicula ; the cotyledons were 

 expanded and of unequal size, by which means the subse- 

 quent relations become intelligible. 



* Ueber den Embryo der Cruciferen. Wiegmann's Archiv, 3ten Jahr- 

 gang, Iten Theil, p. 232. 



