Beer, Development of the poUeii grain and anther of some Onagraceae. 295 



from about 40 to -lö // the protoplast still completely fills the 

 ceU-cavity but it has become entirely free from its walls. 



The further increase in the size of the pollen grain wliich 

 now takes place is more rapid than that of the living protoplast 

 whicli consequently no longer fills the cell - cavity (Fig. 24). 

 We have here, in fact, conditions which strikingly recaU those 

 which Fitting and others ^""*) have described in the case of the 

 megaspores of Isoetes and Selagmella. 



These results have such an important bearing upon our 

 conceptions of the growth of vegetable membranes and render 

 some features of this process so difiicult to understand that 

 several botanists have hesitated to accept them until they coiüd 

 be placed upon a broader basis them was done by those who 

 have examined the megaspores of the Lycopodiales. 



With the exception of Fitting, these authors have exclu- 

 sively rested their conclusions upon microtome sections. Invalu- 

 able as such sections are \ve must not overlook the fact that 

 the long series of manipulations necessary for killing. fixing and 

 embedding in paraffin introduce many possible sources of error 

 and the results obtained by this means should be carefully 

 checked by observations upon living material. 



Fitting worked largely with living spores which he exami- 

 ned partly in physiological salt Solution and ]3artly in water. 



Unfortunalely he gives us no details of liis methods and it 

 would be very desirable to know exactly what was the strength 

 of his physiological salt Solution and whether this particular 

 concentration was found by direct experiment to produce less 

 change in the cell than any other strength. His selection of 

 water as an alternative medium in which to examine the con- 

 dition of the protoplast was most unitable as water is known 

 to affect the protoplasm and its osmotic condition. 



The pollen grains of OenotJiera are particularly favourable 

 for investigation and I have attempted to make my examination 

 of them as complete as possible. Fresh material has been exa- 

 mined in the first place and the results thus obtained have been 

 compared with microtome sections of material lixed with strong 

 and weak Flemming's Solutions, with strong and medium 

 chrom-acetic Solution''), Merke Ts fluid and Worcester's fluid*';. 



1) Fitting, H., „Bau und Entwickelungsgeschichte der Makrosporen 

 von Isoetes und Selagindla etc.-' (Bot. Zeit. Bd. 58. 19Ü0. pp. 107—164.) 



2) Denke, P. , ,.Sporenentwickelung bei ISdaginella'-'-. (Bedielte z. Bot. 

 Centr. Bd. XII. 1902. p. 182.) 



3) Lyon, ;M. F., ,,A study of the Sporangia and Gametopliytes of Sela- 

 ginella Ajnis and .S. Kupestris". (Bot. (iazette Vol. XXXIl. August-Sep- 

 tember 1901. pp. 124—141 and pp. 170—194.) 



■*) Campbell. H. D.. „Studies on the Gametophyte of SelagineUa'-'- . 

 (Annais of Bot. Vol. XVI. 19U2. pp. 419-428.) 



^) Formulae in Chaniberlains „Methods in Plant. Histology' p. 28. 



") Formula for this tluid was obtained from H. S. Beeds paper upon 

 enz3'me secreting cells of Zea and Phoenix. (Ann. Bot. April 1904. p. 271.) 



20* 



