566 Mr. Hassall on the Structure of the Pollen Granule. 



opinion, of easy determination by a reference to the form and 

 structure of the pollen granule. This would place them, as 

 was originally done by L. C. Richard, among monocotyle- 

 dons ; and that this is their true station not the smallest doubt 

 remains in my own mind, notwithstanding that the opinion 

 of most modern botanists appears to lean in the opposite 

 direction. The pollen granule of Nymphcea is oval, hispid; 

 with a furrow down one side, and emits a single pollen tube, 

 thus coming under the definition already given of endogenous 

 pollen. (See fig. 151, 152.) 



Mohl declares that the pollen varies extremely in form not 

 only in genera of the same family, but also in species of the 

 same genus ; and that it even occurs that in some species the 

 anthers contain grains " de formation assez diverses." The 

 two latter assertions, and more especially the last one, are so 

 contrary to the results of my own investigations, and are so 

 opposed to all analogy and to that order and evident design 

 that reign with such constancy throughout all the beautiful 

 works of creation, that I should not have hesitated in confi- 

 dently denying the accuracy of remarks, which would cause 

 such confusion and chaos to hold dominion where nothing 

 but creative skill and wisdom might have been looked for, 

 even before I had personally examined a single granule. 



With regard to Mohl's first statement, it must be admitted 

 that the form and structure of the pollen granule does vary 

 considerably in genera of the same family ; but this is by no 

 means the rule, which should be stated on the other side ; and 

 is, that natural orders, or sections of orders, are characterized 

 by the possession of a pollen granule of one type, and that the 

 more natural and more distinct the affinities of an order, the 

 more frequently w ill the pollen be found characteristic of that 

 order. That it should vary considerably in genera of the 

 same family is nothing more than reason would lead us to 

 anticipate; for it must be allowed that the exact limits of 

 many of our orders are far from being satisfactorily deter- 

 mined, and that some of them contain genera whose true al- 

 liances are far from being clearly ascertained. 



Again, the exact structure of the anthers of some of the 

 genera forming a family will sometimes differ ; and, w<here this 

 is the case, it is only natural to anticipate that a corresponding 

 deviation from the usual form of the pollen granule in that 

 family should accompany such difference. Reference to this 

 fact would frequently account for what otherwise might be 

 ignorantly regarded as a senseless freak of nature, viz. the 

 difference sometimes met with in pollen, the contents of two 

 anthers derived from genera nearly allied, but which anthers 



