542 



II ARSILEACE.C 



In Azolla the chief difficulty I think is presented by the 

 absence of such a developed form of anther as might be 

 expected to accompany so developed a form of pollen grain.* 

 So much so, that each joint of the moniliform filaments, or 

 each grain as found in the ovulum, is not distinguishable 

 from the simplest forms of pollen grains, which I take to be 

 those in which no outer integument is developed, as in Naias, 

 Zanichellia, etc. and which are consequently simple mem- 

 branous bags or sacs. The chieff discrepancy is that the 

 grains of Azolla do not undergo any elongation in the per- 

 formance of their supposed functions as appears to be uni- 

 versally the case in phaenogamous plants, even in those in 

 which the pollen grain is bodily received into the nucleus \% 

 and contrariwise no growth, beyond mere extension, has 



* In order to reduce the usual acotyledonous form of anther to the 

 type of the same organ among phanerogams, I have often speculated on 

 the probability of each anther being a pollen grain. But they have an 

 organic connection with the plants to which they belong, their structure 

 is different, and they generally dechisce. These are all strong objections 

 particularly when it is considered that if these anthers be pollen grains 

 they represent the inner membrane of ordinary pollen. 



But the difference is not unadjustable in my opinion, if the anther of 

 Mosses is compared with the very young phaenogamous anther, at that 

 period when the grume, from which the mass in which the parent cells 

 are developed originates, is so fluid that pressure causes an escape of 

 minute fovillar matter not unlike the contents of the anthers of Musci. 



t I do not mention their organic connection with thfi plant, because 

 that may be the consequence of their not being provided with a proper 

 protecting organ. It is besides at the most only very partial, and it is 

 not greater than that occurring is many forms of supposed anther among 

 these kinds of plants ; i. e. where the anther consists of a cell terminating 

 a stalk of a single row of cellules, to which type the male organ of 

 Azolla and Chara are easily reducible. For any difficulty that might be 

 objected to the attributing fecundating powers to each component cell 

 (not exclusively to the terminal one) becomes lessened by the remarkable 

 form of the male organs of Drepanophyllum and certain Neckerse and 

 Syrhopodon. 



X e. g. Oycas. 



