of the Aquilegia vulgaris. 9 



ened. Lastly, i represents the belly of the cornet, and at k wo, 

 find the gland which, lor its part, secretes whilst the pollen 

 no longer issues from the loculi, and little by little its struc- 

 ture is annihilated. 



This case of metamorphosis not only proves, as we said 

 above, that the two lobes of the cornet of the Columbine are 

 derivations from the cells of the anther, but it puts out of 

 doubt that the tube of the cornet is the lengthened connective. 

 A circumstance which w^e must not lose sight of in this phi- 

 losophic study of a metamorphosed flower, is that the nectar- 

 bearing gland, an organ of emission, and which rids the 

 flower of its excess of carbon, is found at the opposite pole 

 to the pollen-bearing loculi, other organs of emission which 

 also excrete from the individual, but in this case for the pre- 

 servation of the species, a substance eminently charged with 

 carbon. At the two poles then the same function exists, 

 but the one does not begin till the other ceases ; that is to say, 

 the nectarial gland does not exist or become developed until 

 the pollen apparatus wastes away and becomes obsolete. 

 This subject certainly merits a reflection ; even should I be 

 accused of seeing, in Botany, more than my own eyes can 

 see, and especially should I be accused of allowing to myself, 

 in a science of observation, some stretch of imagination. For 

 my part, I could never comprehend how inquiiy into the truths 

 of nature should put aside the understanding, and reduce 

 it to a state of inaction which would render it useless. Be- 

 hind and above facts I always conceive something superior 

 and anterior; for facts are effects, and it is to the know- 

 ledge of causes that we ought to endeavour to come. Now 

 here, in the particular problem which occupies us, I see a 

 verification of the law of organic compensation and a realiza- 

 tion of the unity of composition. Thus, the nectaries are one 

 with the stamens, i\ve stamens one with the leaves, the leaves 

 ONE with themselves, as autochthonous organs. So mvich 

 for the law of unitij. Moreover, the gland is at the end of the 

 nectary, because, by its nature a stamen, the pollen is at 

 the other end; there is a change in the product, but not a 

 change of nature, and by the side of this law oi polarity there 

 is that of compensation ; for, in proportion as the anther-cell 

 closes to render the pollen abortive, there is a development of 

 the gland which begins to secrete the nectar ; the evolution of 

 the gland brings on the atropJiy of the cell, but, fundamentally, 

 there always remains an apparatus of emission. 



Let us return to the Columbine : we have said that one 

 way of transformation, and it is the most common one, w^ould 

 lead to a belief that the nectar-bearing sac may be in some 

 instances the representative of a loculus of the anther. From 



