1887.] NATURAL SCIENCES OF PHILADELPHIA. 331 



noted in any description, except in tlie one case of 0. coccineus in 

 Rothrock's Report of the Wheeler Expedition. Clioisey in DeCaud- 

 olles Prodromes lias "stamina 5" as one of the characters of the whole 

 genus. In view of this uncertainty it is worth placing on record, 

 that 0. hirsutus, Sweet, always within my observations, has five 

 stamens. I have had plants growing from seed gathered by me in 

 Southern Colorado a number of years ago, and the j)lants uniformly 

 have five stamens. 



The flowers always oj^en towards evening. On August 10th, I sat 

 myself to watch their unfolding. At 4 P. M. one could just trace 

 the pink color line, with a portion of a filament discernible ; by 4"30 

 the stamens and jjistil were fully exserted, over an inch in length 

 beyond the perianth, the segments of the perianth were partially ex- 

 panded, but were crumpled, and by 5 these Avere fully expanded, and 

 the flower perfectly in bloom. Hundreds opened, but the anthesis of 

 all was simultaneous. The involucre is three-flowered, but one 

 opens first, and we may call this a central one. The two lateral 

 open together the next day. 



It is interesting to watch the expansion of the flower. The sta^ 

 mens in vernation are incurved. The anther seems to be cauo-ht 

 in the imperfectly open flower. The filament, as it elongates, forms 

 a loop. AVhen the half hour of growth has been nearly reached, 

 tlie perianth by that time expanding a little and, freeing the anther 

 from pressure, the incurved portion springs suddenly forward, and 

 instantaneously we have the long, straight, exserted stamen. In all 

 that I noted the stamens proceeded thus : one at a time successively, 

 the whole five occupying about two minutes to straighten out. 



The flowers continue open all night ; but commence to close by 

 day-light in the morning. 



I have noted on other occasions that some plants do not keep 

 good time any more than some watches. I thought to repeat the 

 observations on the evening of the 11th. Strange to say at 5 P. M. 

 there were no more signs of opening than on the preceding day at 

 4. At 6 P. M. only one flower was open. At 7'30 they were all 

 fully open as the day before at 5. It was a very dark and cloudy 

 afternoon, heavy rain with lightning at 6"30. As most evening 

 flowers seem favored by moisture and darkness, this contrary ob- 

 servation is worth recording. 



The growth of the pistil is coincident with that of the stamens, 

 and is of equal length. The anther cells burst at once on the stamen 



