27 



largest number of flowers of both sexes and this was followed by the 

 Telegraph, \^'hite Spine and Giant Pera. No pistillate flowers were 

 formed on the main shoot in the Telegraph, and only a small number 

 were formed on the main shoot in the Hybrid. The Hybrid was a 

 cross between the White Spine and the Telegraph, having three-fourths 

 or more White Spine stock in it. The tendency to produce almost 

 all of its pistillate flowers on the laterals appears to have been 

 inherited from the Telegraph variety. On the other hand there is 

 practically no difference in the number of staminate flowers on the 

 main shoot and laterals of the Hybrid and Telegraph, whereas the 

 White Spine and Giant Pera show an excess of 5 i per cent of stam- 

 inate flowers on the main shoot. Considering all varieties 52 per 

 cent more staminate flowers were produced than pistillate, and 27 

 per cent more flowers of both sexes were found on the laterals than 

 on the main shoot. There were 93 per cent more pistillate flowers 

 on the laterals than on the main stem. Of the total number of 

 flowers formed in the flrst axils of the laterals 85 per cent were pis- 

 tillate, while the number of staminate flowers in the same axil was 

 15 per cent ; and 91 per cent of the laterals bore pistillate flowers 

 in their first axils. Only 13 per cent of pistillate flowers were found 

 in the second axil of all varieties. The variety which showed the 

 largest number of pistillate flowers in the first axil was the Telegraph. 

 The production of fruit on the main stem in all varieties was, how- 

 ever, constant throughout the life history of the plant. The amount 

 produced on the laterals at first was about the same as on the main 

 shoot and then increased as the plants matured to four and six times 

 as much. There was practically no difference in the time required 

 to mature the fruit on the main shoot and on the laterals in this 

 method of pruning. In these two experiments it is possible to com- 

 pare the differences existing between plants • in which the laterals 

 were pruned and those in which they were not pruned. In the first 

 experiment, where the laterals were not pruned, the yield of 

 fruit was 18 per cent higher on the leader or main shoot than 

 on the laterals; whereas in the latter experiment, where the laterals 

 were pruned, this was reversed, and in this case there were 93 per 

 cent more pistillate flowers formed on the laterals than on the leader. 

 In the first experiment, the fruit on the leader matured earlier than 

 on the laterals ; in the latter experiment, where pruning was prac- 



