304 Belling. 



donianum) that tetrad foniiatioii was nearly iiornial: l)iit that the pollen- 

 grains subsequently shrunk and died, while the enibryo-saes were mostly 

 aborted in an early stai^-e, only one pei'fect enibryo-sae having- been seen. 



Rosenberg (13, 14), in his well-known investigations of the cyto- 

 logy of an apparently sterile hybrid sundew (Drosera longifolia obovata), 

 found that the young pollen-grains mostly lost their contents, while 

 the embrj'o-sacs usually did not develop beyond the liiuiicleate stage; 

 but some ty])ical embryo-sacs were formed, and one embryo resulted 

 from liack-ci'ossing. TISCHLEK (18) investigating the apparently sterile 

 hyliiids resulting from one of C!OHHENs' famous I?ryonia species crosses, 

 found that tetrads were formed, but that the pollen-grains mostly aboi'ted, 

 as did also the embryo-sacs. A few pollen-grains seemed normal. i)ut 

 no perfect embryo-sac was found among tho.se examined. 



(iEERTs (51. in a masterly study of the cytology of the notoiioiis 

 Oenothera lamarvkidiKi of De Vkies, determined that one-half of 

 the pollen-giains, two from each tetrad, aborted, leaving only the cell- 

 wall; and that one-half of the embryo-sacs aborted in the tetrads, the 

 rest being normal. He counted, in sections of owirics. 1 1 1 normal 

 embryo-sacs to 1:^3 aborted tetrads. Most of the s|)('cifs of Onagia and 

 Euoenothera were found by him to have .')<) per cent of aborted pollen- 

 grains, and )H) or .'io per cent of sterile ovules. Oilier related ])lants 

 showed similar comlitions. 'riiougli these Oenotliei'as cannot be consi- 

 dered hybi'ids, yet their lielia\ior in ci'osses is uniipu;. 



OsAWA (1(1) foniid in the Unsliu (Satsiuna) orange, that tetrads 

 were usually formed, but that all except a few of the pollen-gi-aiiis 

 a])orted. Most of the megaspore tetrads also aborted, but a lew normal 

 embryo-sacs were formed, which could be lertili/.ed with jiollen from 

 other varieties of orange. This orange is not known to be a hybrid. 



The hybrids resulting from crosses between Nicotiana tabacum 

 and N. sylvestris have iieen reported l)y several investigators to be (|uite 

 sterile. But HeiJjAIR stated at the genetic conference in Paris in lull 

 that he had obtained progeny from them, and (TtOonsPEKD has matie a 

 thorough investigation (ti), which shows thai a small percentage of 

 viable seeds is formed when certain varieties i>\ .\. tnhacinn are used 

 in the cross. 



It is then evident that the jiartial sterility of some si)ecies 

 hybrids is due in the piiilial iibeiMion of po I le n -gra i ns and 

 emltryo-sacs. The question now is, how this partial sterility is in- 

 herited in the following generations. .Iesenko (7) nuide many polli- 

 nations of wheat by rye. and obtained one hybrid plant with jiollen- 



