THE METHOD OF MOVEMENT 205 



the sieve-tube system. The companion cells, however, 

 practically always fluoresce also. Shortly after appUca- 

 tion of the dye, he observed a marked concentration 

 gradient in the sieve-tube system, grading from strong 

 to faint fluorescence within a few centimeters. He also 

 noted with surprise that the cambiform cells directly 

 adjacent to the sieve tubes were totally dark and remarked 

 on the possibiUty of pecuUar permeability of the sieve-tube 

 walls on these faces. It is interesting to note in this 

 connection that Kienitz-Gerloff (1891) and Fischer (1886) 

 both failed to find plasmodesma connecting cambiform 

 cells with sieve tubes, though between cambiform cells 

 and ordinary parenchyma there are abundant plasmo- 

 desma. Later Kienitz-Gerloff (1902) reports his own 

 observations agreeing with those of Kohl, of Kuhla, and of 

 Strasburger to the effect that in some plants plasmodesma, 

 usually restricted to pits, are observable in the walls 

 between sieve tubes and cambiform and companion cells. 

 It is highly interesting and possibly of considerable sig- 

 nificance that plasmodesma connecting sieve tubes as well 

 as guard cells with their adjacent cells are rare or restricted 

 to small groups in pits. Both in guard cells and in sieve 

 tubes the cells are able to concentrate sugars at concen- 

 trations much in excess of those found in the adjacent 

 cells. Unrestricted and unspecialized protoplasmic con- 

 nections, if highly permeable to sugars, would seem to 

 mitigate against accumulation of sugar or other solutes in 

 one cell over that in adjacent cells. In fact, this very 

 difficulty led me to the finding of the reports of Kienitz- 

 Gerloff on plasmodesma of these specialized cells. Schu- 

 macher attempted to determine whether transfer from cell 

 to cell takes place through plasmodesma. Although in 

 epidermal cells and in sieve plates of Pelargonium he could 

 occasionally see what appeared to be connecting fluorescent 

 strands he could not be certain of it. The strands through 

 sieve plates of Curcurbita, however, were clearly fluorescent 

 and here the strands were obviously the path of movement. 

 When introduced through leaves, with but rare excep- 

 tions even in half-grown leaves, fluorescein moved down- 



