10 M. Morren on the Spur-shaped Nectaries 



a slight examination we should in fact conchide so. As a 

 proof of this, see the states deUneated figures 3, 4, 5 and 6. 

 We often see a stamen, with a filament dilated at its base, 

 take two horns above {c d), whilst one loculus of the anther, 

 inflated, no longer yields pollen ; and the other, being atro- 

 phized to such a degree as no longer to appear except as a 

 yellow gland {b), seems to have produced a rounded sac (e). 

 This sac, the commencement of the cornet, should w-e not 

 suppose it to be a modified anther-cell? and yet w^e have 

 just seen that the tendency of the cells is to produce the 

 lobes of the limb of the cornet, and not its tube. There 

 is a mistake, indeed, as to the true signification of this en- 

 largement, which is nothing but the middle of the connective 

 itself. The connective extends itself outwards, and its hy- 

 pertrophy brings with it the atrophy of the cells or of one 

 cell of the anther; it signifies little whence substance comes 

 to it, so that it only come. This is why the production of 

 the spur does not always cause the whole anther to be meta- 

 morphosed all at once. 



The better-formed cornets, and which even possess all the 

 essential parts, — expanded limb, apex with two lobes and a slit, 

 dilated faux, lengthened tube and terminal gland ; these cornets, 

 1 say, sometimes still exhibit a trace of their old and primi- 

 tive nature in the anther-cell, hardly visible, but distinguished 

 by its yellow colour, whilst all the rest is white and blue, and, 

 above all, distinguished by the grains of pollen that it still 

 encloses in its bosom (fig. 4). 



The conditions (figs. 5 and 6) are tendencies towards a re- 

 gularized form of well-constituted nectaries. The condition 

 (fig. 6) is that found in the common Aquilegice, Nothing here 

 w^ould lead to the supposition of an antherine nature, had not 

 this strange metamorphosis been followed step by step. 

 . It is evident, that aU these cornets being hollow, and de- 

 veloped one above the other in several spirals (fig. 16), all like- 

 wise enter one into another (fig. 8), but it is inexact to say 

 that then the glands no longer secrete. This is a mistake : 

 the secretion continues, and, indeed, the tubes never com- 

 pletely close those into which they have entered. 



Let us now examine in what manner the cornets are gene- 

 rated in a flower of Aquilegia taken at its first periods of de- 

 velopment. 



2. Organogeny of the spur-shaped nectaries. 



To ascertain this organogeny, w^e have followed the method 

 employed by Schleiden and Vogel. Taking a very young 

 bud, which had hardly attained the length of a millimeter and 



