Ida Sutton 283 



Antonowka) though in Lord Derby, for example, such diminution was 

 observed. 



In work of this kind where the measure of sterility or fertility must 

 be the proportion of fruits actually set, it is obvious that many circum- 

 stances must inevitably blur the accuracy of the final result, for in 

 almost all varieties only a proportion, generally small, of flowers can in 

 any case succeed in forming fruits. The allowances to be made for 

 these irregularities must depend largely on personal judgment, expe- 

 rience of each variety, and details which could only be recorded at great 

 length. The classified lists appended are therefore given as representing 

 approximate interpretations of the results recorded in the Tables. 



The stage at which fertilisation fails probably differs in various forms. 

 Mr Collins has made preliminary investigations on this subject. In 

 Coe's Golden Drop, typically self-sterile, he finds that whether the 

 flowers be selfed, crossed, or left unpollinated, the style is definitely cut 

 off by an abscission-layer at its base on about the fifteenth day after 

 the flower opens. The plant's own pollen germinates on the stigmatic 

 surface and tubes penetrate the nutrient stylar tissue, a phenomenon 

 seen by previous observers in other self-sterile plants. 



At the beginning of the work the flowers were bagged on trees in 

 the garden, but from 1912 all experiments were done on trees grown in 

 pots in an orchard-house. This house is provided with mosquito-bars, 

 and various other elaborate precautions were taken to ensure the exclu- 

 sion of insects. It is believed that errors from that disturbance have 

 been almost wholly avoided. 



The numbers of flowers given are in general accurate, but some of 

 these numbers were obtained by estimation based on the number of 

 corymbs used. The word " em " means that the flowers were emascu- 

 lated. In the absence of this expression, the anthers were not removed. 

 All flowers were hand-pollinated. The number of flowers tested is 

 given when a record was made. At first this was not always done. 



The trees, with very few exceptions, were in the finest condition, 

 being for the most part in exhibition state. The horticultural treatment 

 was under the continual supervision of Mr Allard, the Superintendent, 

 and great credit is due to him for the fine health in which the plants 

 were kept, without which such observations would be altogether un- 

 reliable. 



From the horticultural point of view the conclusions to be drawn 

 from our experiments are (1) that many important commercial varieties 

 set little or nothing unless cross-pollinated ; (2) that for the pollination 



