POLLKX CliMX VMilATION 



271 



2. The num})er of veovtativ,- protliallial (vlls certainly varies 

 from one to three, tlioiiuli (lie lati<-r mnnl.rr i. rare. li nia.v als.) 

 lKi[)i)eii that there are none ai all. This (-(.n.lition i^ (■..nii>aral)le 

 with that found in tho.se o-vinnosperins which typically i)ro(lnce 

 no prothalHal cell, as Jimiperus, Taxus, and others. In Picea it 

 was found in 3.8 % of the pollen grains that contained only one 

 antheridium, 



3. The prothallial cells may vary in position as well as in num- 

 ber. The usual arrangement is dorsiventral, forming a single 

 continuous row with the stalk and central cells, but two prothallial 

 cells may be placed laterally in relation to each other instead of 

 dorsiventrally, and in case of three cells the arrangement may be 

 partly lateral and partly dorsiventral (PI. 1, Figs. 5, 0). 



4. In very rare cases the variation in niunber may extend to 

 that part of the row of cells which usually is com})ose.l of the stalk 

 cell and the central cell. An extra cell may be formed at tlii> point 

 perhaps by division from the central cell. The stalk i^ tluMi two 

 cells high. 



5. The niunber ..f prothallial cells formed in the majority of 



ing are to be fully relied upon, since lili..") ', of them showed but 

 one prothallial cell as au-aiiist l.").? which sliowed two. This 

 diflerence is probably greater than it -honld be. mikv it cannot 

 be denied that one cell wa,i hav.- .li.api.caivi in s-)me of those 

 grains whic-h showe.l .>nlv ,.nc. 



^ The membrane, nf tlu' prothallial cclK .wnetlmc. per^i.t 

 is no trice of cellulose w'alls. 



in his Figures () aiul 7. a^ having beeli formed by nn<' instead of 

 two divisions of the pollen mot her-<rll . and tlu- dan-hter .•.-Us 

 havimr been retai.u.l within the .nother-cell wall. In the pollen 

 grai.,. of Pieea >t..<iied by the writer tin. interpretation Mvm. Ncry 



.served arou.id the pollen -rain; .econd, b.van.e of th,' fa. t that 

 the "wings" of the pollen grain were always j.resetu. and always 

 typical in number and arrang<Mnent. ^If these doid)le grain, were 



