14 



and 20//. This mnv be taken as their greatest diameter, 

 for after this tliey woiihl he reckoned part of the condiict- 

 iug tissue. 



The assiniih\tino' cells vary very litth^ iu diameter, heino- 

 about 4-6 X 3-4,a near the apex, and rarely rising aboye 

 8 X 8-4/x in the lowest regions just aboye the basal disc. 

 Their longest diameter is generally parallel to the longi- 

 tudinal axis of the Ayhole vow of cells. 



The increase in thickness of the lower regions of the 

 shoot is brought about not by the addition of thick layers 

 of assimilating tissue, as is the case with PhyUoplioi-a 

 Ih-odid-i. The assimilating layer in Citondi'us is 20-25;x 

 deep in a tlat frond of about 350/x in thickness, but in a 

 frond which was 840/x thick, tlie thickness of the assimi- 

 lating layer was only 25-30/x. The increase in thickness 

 is in fact due to the assimilating cell rows formine- new 

 cells at their tips, whilst their inner cells gradiu\lly pass 

 into the collecting cells, and these gradually pass into the 

 conducting cells. The increase in thickness is noticeable 

 in the central tissue only to any exfent. It is taking place 

 here at the expense of the outer laj-ers, which are, how- 

 eyer, continually being renewed by the formation of new 

 cells at the tips of the rows of assimilating cells. 



It is probable that a good deal of sliding of cells occurs 

 as the growth in length takes place. The increase in 

 length is probably caused not by the central cells actiyely 

 growing in length, but by their being drawn out passiyely 

 during the active lateral extension of the assimilating 

 laj'er. But frecjuent longitudinal slits haye failed to indi- 

 cate in what way tension is distributed in the tissues. 



The central tissue is yery well separated by the 

 filamentous nature of its constituents in the younger parts 

 of the shoot, ])ut in older parts it assumes more and more 

 a pseudoparenchynmtous ap})earance. By this change the 



