LIXXEAX SOCIETI OF LOX DOX, 87 



cells and ovules of a liermaplirodite plaut do not succeed in 

 uuion, as the many self-sterile plants show, rurtherniore, I, till 

 now, could not get seeds by crossing several of my different forms 

 of self-sterile Antirrhinum glutinosum, while I can bv crossing 

 others. 



In all cases of apparent sterility, one must not conclude too 

 soon as to sterility really being absolute. I have for several years 

 tried in vain to make the cross Primula acanlis x ehfior, but am 

 now in the possession of the first generation, I think that iu 

 this case the question is, that certain micro-species of P. acatdis 

 are fertile with certain micro-species of P. ehitior, while others are 

 not. There are more curious cases yet. Miss Digby has shown 

 that Primula jlorihunda and P. verficillata give the hybrid 

 P. kewensis, which produced only thrum flowers and was sterile. 

 It was multiplied by cuttings for about five years, when a single 

 pin-flower appeared on one individual. This pollinated from a 

 thrum flower gave seeds which sprouted ! If we consider, on the 

 other hand, that we now know already a considerable number of 

 fertile species-hybrids — yes, even some fertile bigeneric hybrids— 

 I doubt whether the statement that species generally are infertile 

 after crossing, while varieties are fertile, is so true as it seems, if 

 we exclude from secondai'y difficulties. I believe we must know 

 much moi'e of the real causes of sterility before this statement 

 can be fully accepted. 



Yet it would be very interesting to obtain experimentally from 

 the cross of two self-fertile forms such which were sterile inter se. 

 Now Eosen has already obtained something of the kind. The two 

 closely similar micro-species Erophila stricta and E. elata can be 

 crossed, but the ofi'spring is nearly completely sterile. As in the 

 light of Eosen's experiments E. stricta and E. elata are themselves 

 most probably segregates of a cross, and themselves are quit^e 

 fertile, we have a near approach to the desired proof. Perhaps a 

 case among my segregates from the cross A)itirrhinum glutinosum, 

 and A. majus is better still, yet the evidence is, as the experiment 

 was but once made, so far not perfect. A. glutinosum is self- 

 sterile, A. majus self-fertile. The first generation is self-fertile, 

 in the next segregation both self-fertile and self-sterile forms occur. 

 Now I crossed in the second generation, for another purpose, a 

 self-fertile zygomorphous plant with the pollen of a self-fertile 

 peloric plant. Tjiis was done with ten flowers, and, as I did not 

 doubt the result and other work needed my attention, I did 

 not look at the plant until the time for gathering the fruit came. 

 To my great surprise none of the ten flowers had set fruit, while 

 nearly all other flowers had set beautifully by self-fertihzation. 

 It seeuis rather difilcult to think that this was mere accident, yet 

 of course, the experiment will be repeated. 



One might yet object that crossing is not of sufficient frequency 

 in Nature to allow its playing a considerable role in the formation 

 of species. I think it is fully. According to Kerner, no less 

 than 1000 wild-growing hybrids have been found in the 50 years 



