PBIIS^CIPLES OF FllUIT-G ROWING AKB DEVELOPMENT. 103 



be determined by trial. As an instance, the blossoms of 

 Pyrus Toriiigo — wliicli is a mere bush with fruit not much 

 larger than a pea — were crossed witli pollen of Oldenburg 

 apple. The resulting hybrids were upright and tree-like 

 in habit, with much enlarged leaves, and unexpectedly the 

 fruit yielded enlarged and perfect seed. The size of the 

 fruit was not larger than a small Siberian crab. In this 

 case repeated crossings may give winter apples of real 

 value and of fair size for market. 



107. Gathering Pollen in Advance. — Some American 

 authors advise the use of pollen gathered as needed. In 

 practice this is not easily possible, as the pollen is scattered 

 about as soon as the anthers burst. A more certain plan 

 is to stick a number of small branches loaded with newly 

 opened flowers in moist sand in a warm room. As the 

 anthers become about mature they are clipped off and 

 dropped on white paper. In a dry room the ripening 

 process will soon be completed and the liberated pollen 

 dried. With this dry pollen thus roughly gathered, mixed 

 with the dried anthers and pistils, the most perfect success 

 has been often secured. Even with pollen thus gathered 

 and sent in a letter from Oregon, the success in using has 

 been satisfactory. 



108. Preparing and Pollinating Flowers, — When fruit- 

 tree flowers are fully expanded, or even one half of them 

 have opened, it is not easy to prevent self-pollination. It 

 is far safest to begin the work when the first flowers are 

 nearly expanded. The petals are clippped off with small 

 scissors, followed by carefully clipping off the anthers. 

 At once, after emasculation, the flower or flowers left in a 

 cluster are covered with a paper or muslin bag. The 

 writer's experience favors the use of muslin bags, as they 

 can most rapidly be gathered and pinned to hold the bag 

 in place and to exclude insects. In case high winds and 



