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



rachis and the lower portion of the midrib, these regions at the ex- 

 tremities of the medulla were found to be gradually set off (Figs. 11, 12) 

 from the central portion, or true pith, until in the midrib scarcely a 

 trace of connection can be found between them (Fig. 14). In A. valida 

 there is no such differentiation. The sieve-tubes in these areas are of 

 extreme size as compared with other cells of the plant, the length being 

 too great to measure with a micrometer eye-piece, even when a one-inch 

 objective was used. 



One is led to wonder why these cells remain unbroken while the 

 sieve-tubes of the medullary region break down. It is probable that 

 they are cells specialized for that particular region of the plant, as 

 they appear to be adapted to the conditions they must meet, by their 

 large diameter as well as their length. It may be, however, that the 

 great number of these large tubes makes it possible for them to survive 

 the strain. They may thus be able to retard the growth of the sur- 

 rounding cells by holding them to a slower growth, thus preventing too 

 rapid elongation from snapping them. Owing to the bulging of the 

 walls of the chambers, this region of course grows slightly less rapidly 

 than the bulged region. The relaxation of the central pith tissues by 

 chamber formation probably slightly relieves the tension on these cells. 

 These tubes doubtless serve as conducting channels long after the sieve- 

 tubes of the true pith region break down. 



Since the chamber in the pneumatocyst of Nereocystis would be 

 likely to have a similar origin, a specimen was sectioned. Only mature 

 specimens being available, cross-sections of the side walls of the pneu- 

 matocysts were made. The cells lining the chamber showed signs of 

 lateral strain, much as in mature chambers of A. fistulosa. MacMillan 

 (2) shows the pith area to be surrounded by many sieve-tubes. Within 

 he shows numerous rifts or clefts, diagrammatically. While he advances 

 no explanation of the mechanical cause of this rifting of pith tissues, 

 his observations almost wholly coincide with those herein recorded for 

 A. fistulosa. He states that the sieve-tubes break down before the mature 

 pneumatocyst is formed, speaks of the elongation of the cells of the 

 inner cortex, and says that the earliest form of the pneumatocyst is 

 elliptical, not spherical. The pneumatocyst later becomes globose, chang- 

 ing successively to ovate and pyriform. 



Considering the rifts to be caused, as in A. fistulosa, by the re- 

 sistance of the many sieve-tubes to the rapid elongation of chains of 

 the outer cells, all of these points can be easily explained. Linear 

 strain bulges the sides ; lateral strain aids in producing the globose form 

 on account of continued radial division of the cortical cells ; the sieve- 

 tubes give way, the cortical cells continue to divide, and the resulting 



