554 



The Journal of Heredity 



most important and inclusive, it is 

 difficult to apply a short, descriptive 

 term, though brief discussion will make 

 clear its characteristics. The germina- 

 tion of the pollen grains upon the stigma, 

 the growth of the pollen tube down the 

 style, and generally the union of the 

 proper nuclei within the embryo sac. 

 are apparently normal. Now the 

 interesting fact is, that further develop- 

 mental phenomena proceed to varying 

 degrees. That is to say, the post- 

 fertilization processes may be normal, 

 with a full and complete development of 

 the endosperm and embryo, or there 

 may be practically no development of 

 the egg after its union with the sperm. 

 All gradations exist between these 

 extremes. Very frequently there is 

 practically a complete development of 

 the endosperm but a far less complete 

 development of the embryo. The 

 changes occurring in the ovule are 

 reflected in the external appearances of 

 the young fruits. Time and again 

 there is an apparent set of fruit when 

 blossoms are self-fertilized, but within a 

 period of several days or weeks the 

 young fruit turns yellow and finally 

 drops. On cutting into such fruits to 

 examine the seeds, a highly interesting 

 situation is disclosed. If the fruit 

 has fallen while still very small, it will 

 Ijc noted that there has been very slight 

 development of the ovules, though there 

 is evidence that the endosperm has 

 begun development and often the em- 

 bryo also, but only to a very slight 

 degree. Fruits which have remained on 

 the tree for a greater length of time 

 generally show a much greater degree 

 of development of endosperm and 

 usually, though by no means always, a 

 greater development of the embryo. 

 Fruits which have fallen when half 

 grown generally show what ai)pears 

 to have been a complete development 

 of endosperm and the embryo clearly 

 visiljle to the unaided eye, often filling 

 half and sometimes a much greater 

 proportion of the seed. It is inter- 

 esting to note that in most self-sterile, 

 but not self-fruitful or parthenocarpic, 

 v'aricties of apples and pears, the extent 

 of development of the embryo can be 

 predicted with a high degree of accuracy 



from the observation of the size or 

 weight of the fruit; the larger the fruit 

 the greater the embryo development. 

 This is in accord with the fact already 

 well known that, under like conditions 

 in varieties of apples and pears which 

 are not parthenocarjjic, there is generally 

 an important correlation between weight 

 of seed and weight of fruit. One more 

 condition should be mentioned. In 

 mature fruits, shrivelled seeds, which 

 were apparently of full size before the 

 shrivelling took place, are of common 

 occurrence. Examination of such seeds 

 discloses an entire lack of embryo or one 

 exceedingly small and aborted. Earlier 

 inspection shows the seed-coats filled 

 with endosperm, which shrinks away on 

 maturity of the fruit. 



FAILURE OF EMBRYOS 



How to account for the failure of the 

 embryos at various stages in their 

 development is not entirely clear, but 

 in a study of this problem, some inter- 

 esting phases have arisen. In reference 

 to pollination of orchard fruits, several 

 records are available which state that 

 the variety is self-sterile except for a 

 few misshapen or small fruits harvested 

 from clusters infested with aphids. 

 In our own experiments, we have ob- 

 served this same condition dozens of 

 times. This setting in the presence of 

 aphids has been explained on the 

 grounds that the aphids, in crawling 

 over the blossoms, carried the pollen 

 on to the stigmas, whereas flowers simply 

 inclosed in sacks were not so thoroughly 

 pollinated. The view hardly seems 

 tenable because, unless very hard pressed 

 for time, even when flower clusters are 

 tagged for self-fertility tests, we brush 

 over the stigmas of the flowers with 

 their own pollen; and the most fre- 

 quently occurring condition of this kind 

 with us is that due to the rosy apple 

 aphis which works on the leaves in the 

 cluster, badly deforming them, but, in 

 most circumstances, docs not come into 

 contact with the fruit at all, though the 

 latter is deformed also. In the case of 

 ai)ijle varieties which normally are 

 neither self-fertile nor self-fruitful, in 

 other words self-barren, not only is it 

 possible to cause the fruit to set if 



