322 



MODIFIED CIRCUMXUTATION. CHAP. VII. 



Fig. 126 



us (see also 'Flora,' Oct. 1st, 1873, p. 437) that those ot 



S. napcea fall at night, but 

 to what angle he cannot 

 remember. The leaves of 

 S. rhombifolia and retusa, on 

 the other hand, rise up 

 vertically, and are pressed 

 against the stem. We have 

 therefore here within the 

 same genus, directly op- 

 posite movements. Again, 

 the leaves of S. rJwmbifolia 

 are furnished with a pul- 

 vinus, formed of a mass of 

 small cells destitute of chlo- 

 rophyll, and with their 

 longer axes perpendicular 

 to the axis of the petiole. 

 As measured along this 

 latter line, these cells are 

 only ith of the length of 

 those of the petiole; but 

 instead of being abruptly 

 separated from them (as is 

 usual with the pulvinus in 

 most plants), they graduate 

 into the larger cells, of the 

 petiole. On the other hand, 

 S. napcea, according to Ba- 

 talin, does not possess a 

 pulvinus; and he informs 

 us that a gradation may be 

 traced in the several species 

 S,.da, rhombifolia : circumnutation and of tne genus between these 

 nyctitropic (or sleep) movements of two states of the petiole, 

 a leaf on a young plant, 9 } inches $ida rhombifolia presents 

 high; filament fixed to midrib of ,, ,. .. . ,. < 



newly full-grown leaf, 2| inches in another peculiarity, of which 

 length ; movement traced under a sky- We have seen no other in- 

 light. Apex of leaf 5| inches from stance with leaves that 



sleep : for those on very 

 young plants, though they 

 rise somewhat in the evening, do not go to sleep, as we observed 



