2f)0 Ostenfeld. 



One isolated corymb (two heads) of H. auricula was pollinated with 

 isolated heads of H. auraniiaaim (also from only one corymb). The 

 fruits gathered were sown in April 1908 and produced a family of 

 29 individuals, most of which reached flowering in the autumn of 

 the same year. They presented an astonishing variety or hetero- 

 geneity, and there were not two individuals completely alike and all 

 (with perhaps one exception?) were hybrids. They varied with regard 

 to the colour of the flowers, of which the annexed reproduction of 

 coloured drawings on Plate 4 will give a better idea than long de- 

 scriptions. The specimens in flower in autumn 1908 have been arranged 

 on the plate after flower colour alone, forming a series from H. auricula 

 to H. anrantiacinii. Further, they varied with regard to the size and 

 hairiness of the head; the number of heads; the length and hairiness 

 of the scapes; the form, colour and hairiness of the leaves; the form 

 and vigour of the stolons, e. t. c. No correlation between the variations 

 of the different characters was discovered. 



Some heads of the autumnal flowering specimens were isolated. 

 In 12 specimens all the fruits were empty, but in 5 specimens at 

 least some few fruits were apparently full and were sown in the spring 

 of 1909. Four of these gatherings have germinated and produced a 

 few plants of F„ which, however, owing to the bad summer, did 

 not reach flowering in the same year, and about which I can say 

 nothing more than that the rosettes of each set seem to be homo- 

 geneous. 



In the winter 1908—09, unfortunately, 7 of the 29 plants of Fj 

 died. Among the rest, heads of several individuals were isolated in 

 the summer of 1909 and with similar result as in the preceding year, 

 most of them producing only empty fruits. Some experiments of 

 crossing the hybrids with the parents also gave only empty fruits, 

 but these experiments must necessarily be repeated. 



The hybridization experiments hitherto carried out with H. auricula 

 and H. aurantiacum, thus agree fully with the above-mentioned exten- 

 sive experiments made by Mendel. My experiments have the ad- 

 vantage that only one single corymb of both father and mother has 

 been employed for the cross, so that is cannot be objected that the 

 heterogeneity in Fx depends on different father or mother individuals. 



Thus we have substantiated in the cross H. auricula x aurantiacum 

 an astonishing heterogeneity in F^, and no correlation as to the 

 characters mutually. Most individuals of Fj are sterile, but a few of 

 them bear some few fruits and these are even developed apogamically 



