32 THE CAUSES OP FLUCTUATIONS IN TUKG-ESCENCE 



which they were set. One of them, 0, was now placed in an ammonia- chamber, and 

 the other, b, in a chloroform one. The petals of a immediately began to show purplish 

 black spotting, which rapidly extended over their entire surface. The colour next 

 gradually assumed a greenish tint and ultimately became of a fine warm ochre. 

 Whilst these changes were occurring there were no signs of loss of turgescence, and 

 it was not until some hours after their full completion that a little exudation and a 

 trace of collapse manifested themselves. In the case of b the action of the abnormal 

 medium did not manifest itself for some time; but, when once initiated, alteration in 

 colour and loss of turgescence ran hand in hand, and after the close of a few hours 

 the colour was of a dull maroon red and the tissues of the flower were completely 

 flaccid and collapsed. Twenty-four hours after the initiation of the experiment, a was 

 slightly collapsed, but after this it remained apparently unaltered for some days. 



Experiment XXII. — Two flowers of scarlet Hibiscus were taken, and one, «, 

 suspended free, apex downwards, in an ammonia- chamber, whilst the other, b, was 

 similarly hung in a chloroform-chamber. In the case of «, discolouration, in the form of 

 black spotting of the petals, made its appearance at once, and within an hour the red 

 had entirely gone, save in a few isolated patches, and the rest of the tissue was deep 

 purplish-black, passing into deep greenish and ochre at the margins. There was, how- 

 ever, no indication of any tendency to collapse. In b discolouration was developed 

 much more slowly; but at the close of an hour the original vivid scarlet had 

 been replaced by a deep maroon, and the petals, which had originally been highly 

 reflexed, were drooping downwards and rapidly collapsing. The first traces of collapse 

 in a did not appear until three hours after the beginning of the experiment, and even 

 then the petals remained highly reflexed. On the following day the petals of a were 



drooped, but still widely divergent and firm in texture (plate III. fig. 3); whilst those 

 in b hung vertically downwards and were perfectly flaccid (plate III. fig. 4). Twenty- 

 four hours later the petals of a were slightly less divergent than they had been, but 

 were still so divergent as to allow the stigmas to be visible beyond them in profile. 

 Both flowers were now removed from the chambers and hung in the open laboratory, 

 where they remained for days apparently unaltered, b being fully collapsed and a 

 retaining a considerable amount of turgidity. 



Experiment XXIII, — Two flowers of scarlet Hibiscus were, as in the previous 

 case, placed respectively in ammonia and chloroform-chambers. They were not, how- 

 ever, suspended, but were set with the freshly, subaqueously divided extremities of 

 their stalks immersed in water. As in the previous experiment, the flower in 

 the ammonia chamber began to show black spotting immediately, and rapidly passed 

 on through stages of deep purplish black and deep greenish to a uniform warm ochre 

 without showing any signs of collapse; whilst that in the chloroform-chamber took 

 some time to show any signs of change in colour, but when it once had begun to 

 discolour, began also to collapse rapidly. On the following morning a slight amount of 

 collapse only was evident in a, whilst b was completely collapsed, flaccid, and soaked. 

 The ammonia- chamber was allowed to stand unopened for a week, and during 

 that period a remained almost in the same condition as it was in after twenty-four 

 hours' exposure. 



From the results obtained in the first of these experiments it is evident that the 

 retention of turgescence in such cases is not wholly due to the presence of free 



