48 Puget Sound Marine Sta. Pub. Vol. 1, No. 8 



the others having broken down. The ends of the rifts are sharply 

 angled, as if torn (Fig. 4). The walls of the rift are lined by many torn 

 hyphal chains. 



The first true chamber differs comparatively little from these well- 

 formed rifts. The walls have fewer chains of hyphal cells, and the ends 

 of the cavity are less sharply angled (Fig. 13). The septa are more 

 compact, but the tissues along the sides of the chamber are rather loosely 

 arranged, thus allowing for later enlargement of the septa as the midrib 

 grows. 



The cross-section (Fig. 5) now shows a solitary rift, or line of 

 cleavage, parallel to the long diameter of the pith. The chamber thus 

 formed has the ragged chains of hyphal cells running inward from 

 its walls, as noted in the longitudinal section. The cells of the inner 

 cortex show increased strain, as if being stretched by some mechanical 

 means. 



In the nearly mature septum (Fig. 20) the cells of the inner cortex 

 so closely resemble those of the septum that no sharply defined line of 

 demarcation now appears between these tissues. The walls of mature 

 chambers are practically free from hyphae. The extremities of the 

 chambers are no longer acutely angled, but the ends are curved outward, 

 or may be almost flat (Fig, 17). The septa appear to remain about 

 the same thickness throughout the midrib. Sloughing off of broken 

 down cells about balances the increase in thickness as a result of cell 

 enlargement. 



In cross-section the acute angles at the extremities of the rift or 

 chamber (Fig. 16) are seen to become less acute, and in the mature 

 chamber (Fig. 14) the angles have disappeared, leaving the chamber 

 oval in cross section. The tissues have been strengthened at the ends 

 of the rift to prevent further tearing. The inner cortical cells still show 

 the strained appearance already referred to, but the strain seems 

 lessened. 



There seems to be no definite length of chamber, but there is a 

 gradual increase in chamber length from the base of the lamina to its 

 widest part, though the lengths vary widely. The chambers in a single 

 plant were measured in millimeters throughout this region, from below 

 upward, the results being as follows: 8, 11, 9, 6, 11 25, 9, 8, 5, 13, 11, 

 17, IS, 19, 31, 20, 27, 17, 11, 2, 20, 23, 23, 19, 19, 38, 37, 34, 17, 42, 

 36, 42, 12, 47, 50, 5Q, 73, 69, 56, 58, 94, 56, 138, 65, 117, 113, 97, 75, 

 100, 83, 120, 130, 156, 103, 67, 101, 103, 128, 153, 131, 142, 181, 150, 

 70, 169. 



The elongation of the sieve-tubes by the continued division of the 

 cortical cells after the sieve-tube cells cease dividing, forms the basis of 



