66o Journal of Agricultural Research voi. xii, no. lo 



I have already pointed out that up to the time of liberation of the 

 microspores from the tetrad, in the strawberry, cell divisions have 

 resulted merely in an equal division of the cytoplasm of the original 

 mother cell between its four granddaughter cells with no evidence of 

 any metabolic changes resulting in an increase of cytoplasm. A similar 

 condition exists in those forms which show irregular divisions during 

 reduction with uneven distribution of cytoplasm between the resulting 

 cells. Liberation of the microspores from the tetrad marks the end of 

 the period of dependence of these cells upon the 2X generation, as far 

 as future growth and development is concerned. At this time there is 

 no difiference between the conditions surrounding the microspores of a 

 hybrid and those of a genetically pure individual. Both groups of 

 microspores are set free in a homogeneous anther sap to complete their 

 own further development. 



The progress of the developmental stages in plants of pure F. 

 virginiana is characterized by great regularity of development of the 

 individual microspores within a loculus with regard to rate of enlarge- 

 ment, time of division of the microspore nucleus, and the subsequent 

 development of cytoplasm. In fact, there is as great uniformity shown 

 in these stages as was shown in the stages leading up to the mature 

 tetrad stage in either F. virginiana or Miimesota 3. Minnesota 3, on 

 the other hand, shows great irregularities during this developmental 

 period in rate of growth of the individual microspores, in the time 

 of division of the microspore nucleus, and in the rapidity of formation 

 of cytoplasm. This lack of uniformity is in striking contrast to the 

 uniformity shown in earlier stages of the same plant, while the cells 

 were dependent on the 2X individual. The liberated microspores of 

 Minnesota 3 are strikingly uniform in size and cell contents. 



During any period, following liberation of the microspore to the 

 completion of development, microspores or pollen grains may be found 

 degenerating. As all of the grains within a loculus are free in a 

 homogeneous nutrient liquid, it seems difficult to believe that the varia- 

 tions in development can be due to anything but the individual 

 constitution of the microspores. 



Indeed, there is constantly accumulating an increasing amount of 

 evidence which points to the continual elimination of gametes bearing 

 certain chromosome combinations. In 1894 Millardet (27) reported on 

 a series of hybrids between species of strawberry the progeny of which, 

 he said, formed an exception to the general rule of hybrids, as the specii&c 

 type of one or the other parent was always produced in the first and later 

 generations. The specific type shown in the second generation was, 

 with one exception, the same as that shown by its parent in the F^ 

 generation. Millardet mentions complete sterility in one species of cross 

 and high percentages of sterility in the F^ generation of some of the other 

 combinations. Bellair (j) reports that in the tobacco cross Nicotiana 



