LIFE HISTORY OF PINUS 1 35 



cells are not formed ; the smaller sperm-nucleus is never in ad- 

 vance ; the generative cell does not give rise to the binucleated 

 sperm-cell until after the stalk-cell has passed beyond it, nor has 

 its normal division ever been observed to occur while it is still 

 united with the pollen-grain by its own cytoplasm. In this case 

 it seems very evident, then, that two generative cells have arisen 

 by the division of the first generative cell. Whether both of 

 these would have divided to produce four sperm-nuclei is of 

 course a mere matter of conjecture, but the cytoplasm of the 

 smaller cell is very scanty and it is probable that only the larger 

 one would have functioned as the generator of the sperm-nuclei. 

 The uncertainty as to the origin and fate of these extra nuclei is 

 in each instance too obscure to admit of any theorizing regard- 

 ing their significance. 



Usual Conditions in the Female Ganietofhyte. — In only one in- 

 stance has more than one macrospore-mother-cell been observed. 

 In this case two cells which are very similar and centrally placed 

 in the spongy tissue differ from the surrounding cells in exactly 

 the same way as has been described for the young macrospore- 

 mother-cell (fig. 260, plate XXIII). Farmer ('92) records the 

 discovery of a double prothallium in Pinus sylvestris^ and Hof- 

 meister had previously made a like observation in the same 

 species. I find no other instance recorded for Pimis in w^hich 

 more than one macrospore must have been functional. Shaw 

 ('98) and Arnoldi ('99) find one or more macrospore-mother-cells 

 in Sequoia from which several embryo-sacs may arise ; Arnoldi 

 ('00) has made a similar observation for Cunni^ighamia , Sciad- 

 Q-pitys^ Taxodium, and Cryptomeria ; Lotsy ('99^) and others 

 find many young embryo-sacs in Gnetum; and Coker reports 

 the prescence of two prothallia in Prodocarfus and Taxodium. 

 The presence of a multicellular sporogenous tissue has been 

 reported in the Angiosperms by several students — Nawaschin 

 ('99^) in Corylus^ Lloyd ('01) in the RitbiacecB^ Murbeck ('01) 

 in the PosacecB, and by others. The appearance of more than 

 one functional spore within the ovule of such widely-separated 

 plants makes it rather doubtful if this character is important 

 phylogenetically. 



Juel ('00) found that the walls separating the macrospores in 



