154! THE STRUCTURE OF FLOWERS. 



together or aggregated, so as to occupy only two-thirds of 

 the space they formerly required. The contents, retreating 

 from different points of the circumference, are condensed in 

 the centre of the cell, and the transverse stri^ are pro- 

 nounced in a hiofh desfree. The cells at the back of the fila- 

 nient are contracted in repose, and extended under irritation, 

 i.e. in an opposite manner to that of the other side of the 

 filament. The irritability does not reside in the epidermis. 



A I'esult of this aggregation must be a frequent displace- 

 ment of the nucleus. In Weiss's figure the irritation hap- 

 pened to be made apparently at one end of the cell, while the 

 nucleus was at the other ; but in Heckel's description it 

 appears that the protoplasm is dra^vn from every point ; 

 so that, supposing the nucleus had been at the lower end 

 of the cell (Fig. 45, a), it would have been most probably 

 displaced. The consequence would be, tbat if such a nucleus 

 formed its cell-plate, the ultimate position of that plate 

 would be different from Avhat it Avould have been had no 

 irritation been applied to the organ. 



Though one does not look to electricity as a cause in 

 nature, yet that liglit determines the direction of cell-division 

 is abundantly proved in the case of leaves, whose tissues alter 

 according to their position ; the palisade cells, for instance, 

 bring formed on both sides, if the exposure to light be equal, 

 or on the under side if that be placed uppermost. Similai'ly 

 does it influence the formation of stomata.* Again, Stahl has 

 shown that the direction of the division of the nucleus, 

 which takes place in the spores of Equisetum depends upon 

 the direction of the rays of light ; the two daughter-nuclei 

 lying in the direction of the ray. On the other hand, the 

 nucleus at the greater distance from the source of light is 

 that of the root-cell, while the one nearer to the source of 

 * M. L. Dufour., Ann. Sci. Nat., torn. 50 (1887), p. 33 1. See below, p. 173. 



