114 TORREYA 



the end of the night, and before dawn (about 5:15 a.m. at this season) they 

 are already slightly spread. The petals bend outward very slowly, and it may 

 be two hours after they first begin to expand (or 7:15 a.m.) before they 

 reach their fullest opening. The flowers are evidently functional, however, 

 long before this. The fresh stigmas are sufficiently exposed to receive pollen 

 when the flowers are half open, an hour after the sky begins to brighten. The 

 nectaries begin to secrete their sweet fluid as soon as the sun falls upon the 

 trees, and insects come to the flower. 



Between Ay 2 and 5 hours after daybreak, or from about 9:45 to 10:15 

 a.m., according to the day and the individual tree, the flowers begin to close, 

 with their anthers still tightly sealed and their pollen all unshed, but with 

 their stigmas as a rule somewhat discolored. In from forty-five minutes to an 

 hour after they display the first signs of closing, the flowers have all practi- 

 cally closed, although the final tight pressing together of the edges of the 

 petals, until the once-opened flowers come to resemble larger buds (Fig. 3), 

 takes place more slowly. Then there are always a few laggards — flowers which 

 have been injured by the bites of insects, or are otherwise not in perfect condi- 

 tion — that fold up more slowly. 



These once-opened flowers remain tightly closed until they begin to open 

 on the following day at about the same time that they began to close after their 

 first anthesis. They require about three-quarters of an hour to reach full ex- 

 pansion. On some trees, the anther-valves begin to lift and to expose the pollen 

 before the petals have fully spread ; but on others they open more tardily, 

 and do not begin to rise until half an hour or more after the flower has com- 

 pleted its expansion. Simultaneously with the opening of these second-period 

 flowers, another set of flowers on the same tree, which opened for the first 

 time that same morning, are closing again. Soon after two o'clock in the 

 afternoon, the flowers, which have shed all of their pollen, close for the second 

 time, after an opening which lasted between three and four hours. By three 

 o'clock or a few minutes later, if the afternoon is dry and warm, as it usually is 

 at this season, all but a few laggards have closed (Fig. 4). On wet afternoons 

 the closing of the flowers may be considerably delayed. Once folded up, they 

 do not open again. Those which fail to be pollinated, or at least fail to be 

 fertilized, fall in their entirety after lingering on the tree for a few days, 

 during which they remain tightly closed. Those whose ovaries begin to develop 

 into fruits — apparently always a small minority — shed only the perianth and 

 the stamens. 



If the flowers of all of the aguacatillo trees in any neighborhood behaved 

 exactly as those whose history has been followed, it would be very difficult 

 for them to be pollinated ; for at the time their stigmas are receptive there 

 would be no pollen shed, and while they are shedding their pollen, there would 

 be no stigmas in condition to receive it. A necessary part of this complex 

 mechanism is a second class of trees, Class B, whose floral behavior is com- 



