89 



closed two nights and a day, and expand again on the second 

 morning following. 



During the morning, bees and other insects may carry pollen 

 from Class B trees with flowers in their second opening to the re- 

 ceptive stigmas of Class A flowers in their first opening, while 

 during the afternoon the process is reversed. 



All trees of the same horticultural variety belong to the same 

 class and their flowers open and close, under like weather condi- 

 tions, at about the same time. Indeed, since they are all derived by 

 vegetative means of propagation from the same tree, they behave 

 exactly as though they were branches of a single individual. 



Stout in California and Florida and Robinson and Savage in 

 Florida have determined the classification of a large number of 

 named varieties. 



The small, yellowish flowers of the avocado are borne in pani- 

 cled clusters on short, leafless branches at the base of the annual 

 shoot (Fig. 1 ) . The fully expanded blossom measures about 12 mm. 

 in diameter. The six perianth divisions are alike in size and color. 

 There are nine stamens arranged in two rows, the six of the outer 

 series situated opposite the perianth divisions, the three of the 

 inner series inserted just inside of those three of the outer series 

 which lie opposite the outer perianth divisions or sepals. In the 

 center of the flower is a single one-ovuled pistil with short style 

 and simple, capitate stigma. During the first opening, when the 

 stigma is receptive, the nine stamens with their closed anthers are 

 bent outward and lie above the inner face of the perianth divisions, 

 leaving the stigma fully exposed (2, Fig. 2) . At the second opening, 

 when the stigma is usually discolored, the three inner stamens, 

 their filaments somewhat elongated, stand upright in the center of 

 the flower overshadowing the pistil, while the other six curve 

 obliquely upward (4, Fig. 2). Each anther opens by four neat, up- 

 lifted, elliptical valves and sheds its yellowish pollen. 



After I had become familiar with the general type of behavior 

 of the trees by preliminary observations, I devoted February 16 

 and 17, 1929, both clear, warm days, to the observation of eight 

 trees whose lower branches were accessible from the ground or 

 from an eight-foot step-ladder. On each of six of these trees twen- 

 ty flowers were marked with small jeweller's tags, so there could 

 be no doubt of the behavior of individual blossoms. Both days, 

 from 7:30 a.m., when the earliest flowers were opening, until sun- 



