288 MODIFIED CIKCUMNUTATION. Chap. VI, 



and in tlie morning two of them were dead, wMlst not a single 

 other leaf on the many plants was even injured. 



On the 23rd the pot was exposed for 1 h. 30 ra., the tempera- 

 ture on the grass being only — 2° C, and not one leaf waa 

 injured: the pinned open leaves, however, all stood from 

 5 to I of an inch above the cork. 



On the 24th the pot was again placed on the ground and 

 exposed to a clear sky for between 35 m. and 40 m. By a mis- 

 take the thermometer was left on an adjoining sun-dial 3 feet 

 high, instead of being placed on the grass ; it recorded 25° to 

 26° F. (- 3-3° to - 3-8° C), but when looked at after 1 h. had 

 fallen to 22° P. (— 5'5° 0.); so that the pot was perhaps exposed 

 to rather a lower temperature than on the two first occasions. 

 Eight leaves had been pinned out, some close to the cork and 

 some above it, and on the following morning five of them (i.e. 

 63 per cent.) were found killed. By counting a portion of the 

 leaves we estimated that about 250 had been allowed to go to 

 sleep, and of these about 20 were killed (i.e. only 8 per cent.), 

 and about 30 injured. 



Considering these oases, there can be no doubt that the 

 leaves of this Oxalis, when allowed to assume their normal 

 vertically dependent position at night, suffer much less from 

 frost than those (23 in number) which had their upper surfaces 

 exposed to the zenith. 



Oxalis carnosa. — A plant of this Chilian species was exposed 

 for 30 m. to a clear sky, the thermometer on the grass standing 

 at — 2° C , with some of its leaves pinned open, and not one leaf 

 on the whole bushy plant was in the least injui'ed. On the 

 16th of March another plant was similarly exposed for 30 m., 

 when the temperature on the grass was only a little lower, viz., 

 - 3° to - 4° C. Six of the leaves had been pinned open, and 

 next morning five of them were found much browned. The 

 plant was a large one, and none (>{ the free leaves, which 

 were asleep and depended vertically, were browned, excepting 

 four very young ones. But three other leaves, though not 

 browned, were in a rather flaccid condition, and retained their 

 nocturnal position during the whole of the following day. In 

 this case it was obvious that the leaves which were exposed hori- 

 zontally to the zenith suffered most. This same pot was after- 

 wards exposed for 35-40 m. on a slightly colder night, and 

 every loaf, both the pinned open and the free ones, was killed. 

 It may be added that two pols of condculata (var. Atro- 



