296 MODIFIED CIRGUMNUTATION. Chap. VL 



a little above the cork. Again, the tips of some few 

 leaflets, which had been pinned close to the cork, pro- 

 ected a little beyond the edge, so that the air could 

 circulate freely round them. This occurred with six 

 leaflets of Oxalis acetosella, and their tips certainly 

 suffered rather less than the rest of the same leaflets ; 

 for on the following morning they were still slightly 

 green. The same result followed, even still more 

 clearly, in two cases with leaflets of Melilotus officinalis 

 which projected a little beyond the cork ; and in two 

 other cases some leaflets which were jjinned close to 

 the cork were injured, whilst other free leaflets on 

 the same leaves, which had not space to rotate and 

 assume their proper vertical position, were not at all 

 injured. 



Another analogous fact deserves notice : we observed 

 oh several occasions that a greater number of free 

 leaves were injured on the branches which had been 

 kept motionless by some of their leaves having been 

 pinned to the corks, than on the other branches. This 

 was conspicuously the case with those of Melilotus 

 Petitjnen-eana, but the injured leaves in this instance 

 were not actually counted. A^'ith Arachis lujpogxa, a 

 young plant with 7 stems bore 22 free leaves, and of 

 these 5 were injured by the frost, all of which were on 

 two stems, bearing four leaves pinned to the cork- 

 supports. With Oxalis earnosa, 7 free leaves were 

 injured, and every one of them belonged to a cluster 

 of leaves, some of which had been pinned to the cork. 

 We could account for these cases only by supposing 

 that the branches which were quite free had been 

 slightly waved about by the wind, and that theii 

 leaves had thus been a little warmed by the sur- 

 rounding warmer air. If we hold our hands motion 

 less before a hot fire, and then wave them about, we 



