74 



MORPHOLOGY OF AXGIOSPERMS 



B 



Fm. 30. — A, Fatsia japonica. Longitudinal 

 section of nucellus showing two tetrads; 

 xl75. B, similar section of Aralia race- 

 ?nosa, in which the nucellus bears a strong 

 resemblance to a microsporangium ; x 340 

 —After DccAMP. iia 



pollen mother-cells of Typha (Schaffner 49 ) (Fig. 57) the 

 microspores are formed in rows of four as well as tetrahedrally, 

 while in Asclepias (Strasburger, 96 Frye") the microspores 



constantly appear in rows of 

 four (Fig. 58); and in the 

 pollen mother-cells of Zos- 

 tera (Rosenberg 93 ) the four 

 elongated microspores lie 

 side by side in the same 

 plane. Xor is it a criterion 

 of a tetrad that all of its 

 spores shall mature, for in 

 Care.v acuta (Juel 88 ) (Fig. 

 59) only one spore of the 

 tetrad functions, a habit to 

 be observed also among the 

 Pteridophytes, as in Mar- 

 silea and Azolla. Even the rare case of more than four mega- 

 spores in a row is met by the occasional occurrence of more than 

 four microspores in the pollen mother-cells of Hemerocallis fulva 

 (Juel, 50 Fullmer 65 ) and of Euphorbia corollata (Miss Lyon 54 ) 

 (Fig. 60). The usual tetrad arrangement, however, is not lack- 

 ing among megaspores, as observed by Ducamp 112 in Fatsia 

 japonica (Fig. 30), in which after the mother-cell had divided 

 transversely the two daughter-cells were observed to divide longi- 

 tudinally ; and in one preparation, in which two mother-cells had 

 thus divided, the nucellus looked very much like an ordinary spo- 

 rangium. In another case the middle cell of a row of three had 

 divided longitudinally. The parallel seems still more striking, 

 when microspores germinate like megaspores, even reaching the 

 stage with eight free nuclei, as observed by Xemee 5S in the petal- 

 oid anthers of Hyacinthus orientalis, whose microspores some- 

 times show three successive mitoses, giving rise to four nuclei at 

 each end of the pollen-groin (Fig. 31). Even the formation of 

 three cells at one end, and the wandering of one polar nucleus 

 toward the middle were observed, although fusion did not occur. 

 Nemec did not hesitate to homologize these divisions with those 

 occurring in the embryo-sac. There seems to be no longer any 

 reasonable objection to the view that this row of cells, whose 

 formation is initiated bv the reduction division, is the homo- 



