206 



The Journal of Heredity 



^ 



# 





^i^ 



^ 



that self-sterility in plants may be 

 analogous to wheat which is immune to 

 the rust fungus. The rust fungus is an 

 obligate i)arasite and when the sjiore 

 germinates on the host the mycelium 

 ])enetrates the tissue. In the immune 

 l)lants the tissue is immediately killed. 

 Since the fungus is an obligate parasite, 

 it cannot grow in the dead tissue of the 

 host and so dies. Thus the plants which 

 are immime are ver\' susceptible to the 

 fungus. \\'hen this analogy is apjjlied 

 to self-sterility, it ma\' mean that the 

 pollen tube from a self-iK)llination is in a 

 more favorable mediiun for food assim- 

 ilation. Since the ]jrimary function of 

 the pollen tube is to assimilate food, it 

 does not continue to grow at so fast a 

 rate as a pollen tube from a cross- 

 ]:)ollination where the medium is not so 

 suital)le to it. Tlie i^ollen tube from a 

 cross-])ollination is assuined to continue 

 growth in order to obtain a better food 

 su])ply. 



It is possible to interpret the greater 

 width of the pollen tubes from a self- 

 pollination of Tradescantia in two ways. 

 First, there may be an inhibitor either 

 in the pollen grain or stigma which in- 

 hibits the growth of the pollen tube from 

 a self-]:)ollination ; and since the pollen 

 tube cannot lengthen itself, it grows 

 wider. Or. second, the food sup])ly 

 may be more favorable and so the 

 tube has not the incentive to prolong 

 itself. 



A LONG Tl IIK 



While the above ijollcn tube is lonjj as 

 compared with the one l)csiilc it, 

 whieh i.s just l)eKinninK to germinate, 

 it has liy no means reaehed tlie max- 

 imum length. In maize, the j)ollen 

 grain falls fjn the end of the ".silk" and 

 must >jrovv all the way down that to 

 the seed, a distanee of a foot or more 

 in many eases. When it is remem- 

 bered that the ()ollen grain is not mueh 

 larger than a i)article of dust, and 

 sometimes ean hardly be seen without 

 a mieroscopc, its great ea|)acity for 

 growth will be appreeiatcd. Photo- 

 micrograph by John Howard I'ainc. 

 (FiK. 4.) 



CORREXS HYPOTHESIS 



Taking the first alternative, Com])ton 

 (1913, ]). 203) has shown that the in- 

 hibitor h>pothesis as put forth by Cor- 

 rens (1912) does not exjjlain the phenom- 

 enon of self-sterility. Correns worked 

 with Cardaminc pratensis, and he as- 

 sumed that self-sterilit>' was due to two 

 allelomori)hic inhibitors, the jjresence of 

 either being able to cause self-sterility. 

 Correns made crcjsses between plants 

 which, according to his hypothesis, 

 should have given him fotir classes of 

 ])rogeny, one of which should lack both 

 inhibitors and so be self-fertile. How- 

 ever, all of the i)lants were self-sterile, 

 so that the experimental evidence ob- 

 tained by Correns did not ctMifirm the 

 liN'pothesis. 



East (1915, p. S5j suggests that the 



