THE FLOWKU BKHAVlOll OL^ AVOCADOS 



167 



It should be noted that the tented trees received a highly 

 forced pollination. Day ai'ter day the bees worked most indus- 

 triously among the flowers. On most days every flower was 

 probably visited many times. Obviously flowers in the second 

 opening had pollen liberally distributed over the pistils as soon 

 as pollen was shed. Whenever there was an overlap the j^istils 

 of flowers in the first opening had an excellent chance to be close- 

 pollinated not once as in hand pollinations, but many times. For 

 Linda and Panchoy second-period flowers normally shed pollen 

 in the forenoon and possibly bees may have carried pollen on 

 their bodies from late forenoon until the time in the afternoon 

 when first-period flowers were open. For the Taft there would 

 be no such chance as this but only direct and almost immediate 

 self-pollination of second-period flowers or of close-pollination 

 when there was overlap, which the season's record (chart 9) 

 shows rarely occurred in 1925. 



For the Trapjj with its first-period flowers opening late in the 

 afternoon (see charts 3, 4, and 10) it would seem that the best 

 chance for fruit-setting when tented w^ould be to self-pollination 

 of flowers that had opened late on the afternoon of the previous 

 day or had skipped this period of opening as is frequent for the 

 variety. 



It would seem also from flower behavior that there is very 

 little oi)portunity for the first-period flowers of Trapp to be 

 cross-pollinated. They open so late that there is scarcely any 

 pollen available from A varieties. On account of this the chances 

 for fruit-setting in the Trapp may be fully as good for a tented 

 tree as for trees in the orchard, or even much better when the 

 tented tree is given a hive of bees. 



On the basis of flower behavior one might predict that the 

 flowers of Taft with their 36-hour cycle (.4 in chart 5) and 

 scarcely any overlap (see chart 9) would set very few fruits 

 without cross-pollination and also that Trapj) with a short cycle 

 and frequent skipping of the first opening would be most liable 

 to set fruit to self-pollination accomplished during the second 

 period of flower opening. 



There is no question that the flowers of the tented trees re- 

 ceived bee visitations far in excess of what flowers of any 



