348 MORPHOLOGY. 



exists. Coherence of every kind occurs here ; the stamens some- 

 times become blended in the entire length ; or the filaments cohere 

 in part or entirely ; the filament with the perianth or the corolla. 

 Coherence of their stipules also occurs, as in the Amarantacece. 



"We have here again a few points to bring forward, which require a 

 more minute exposition in order to establish an accurate comprehension 

 of the stamen. 



In the first place, I must discuss the proper definition of the stamen. 

 In this matter I need spend no more words than are necessary to say 

 that it is a modified leaf, since all Botanists whose opinion is of the 

 slightest consequence are now agreed upon this point : but this does 

 not do much for the formation of the definition ; we have such a multi- 

 tude of kinds of foliar organs, which comprehend the whole region of 

 possibility of conditions of position, form, colour, and structure, that it is 

 necessary at once to draw a line between the stamens and all other forms. 

 As a foliar organ of the flower, its definition is not determined, since 

 the sphere is infinitely great here. According to the principle which I 

 have placed at the very beginning of the whole study, namely, the mor- 

 phological mode of treatment, there are only two possible ways of defin- 

 ing accurately, viz. according to the external and internal forms, or ac- 

 cording to the condition of structure. According to the outward form, the 

 externally visible anther-cell, and according to the structure, the develop- 

 ment of the pollen, are undoubtedly the characteristics which define the 

 stamen as such : both are so intimately connected that it is unimportant 

 which character is taken. If this character be passed over, scarcely 

 any stamen can be distinguished from the accessory foliar organs of 

 the flower ; many for example, the outer stamens of Nymph&a, the 

 stamens of Canna not in any way from petals, &c. And, therefore, 

 the definition must be thus taken : a stamen is a foliar organ of the flower 

 which develops anther-cells, and contains within these pollen. By such a 

 definition we acquire a safe basis for the comprehension of the flower, 

 and the accurate description of the forms. Nothing which does not 

 correspond to this definition (and no other is possible) is a stamen. On 

 such grounds, therefore, it is altogether incorrect and superfluous to 

 speak of castrated or abortive stamens, i. e. of stamens which are not 

 stamens at all. An imperfect perception of the nature of the flower as 

 a whole, lies at the bottom of such expressions. This consists of foliar 

 organs (and axial organs) variously modified, some of which must 

 be stamens (or seed-buds), or the definition of a flower cannot be 

 retained. But the essential nature of the flower does not by any means 

 determine how many foliar organs shall be developed into stamens. 

 Even in particular groups of plants no law can be deduced, seeing that 

 nature forms sometimes one way and sometimes another ; but what lies at 

 the basis of the groups, as types, are definite conditions of development, 

 through which are conditioned the number and arrangement of the foliar 

 organs, but not particular modifications of them. These latter are, 

 perhaps, of quite subordinate importance, and may alter in genera and 

 species, nay, even in mere varieties, sportive forms, and monstrosities. 

 What I have particularly to do here, is to reject the anthropomorphic 

 preconceptions of certain ideal types, which float between us and nature, 

 and sometimes perfectly, sometimes imperfectly, attain to a likeness of 

 her ; which, however, we entirely carry over into nature, instead of 

 obtaining from her, and which at best can serve but as a make-shift, 



