334 Pub. Paget Sound Biol. Sta. Vol. 2, No. 53 



of the midrib, trace its lateral branches out to the edge of the cortex, and 

 note their continuation, through the short cells of the intercalary growing 

 point, into the axis of the assimilating hairs. Just at the edge of the 

 cortex, the basal cell of the assimilating hair and the outer cell of the 

 midrib branch round away from each other at their separating wall, mak- 

 ing the axis thinner at this point and therefore weaker. The hair breaks 

 off at this weak point and the cortical cells, by lateral divisions, soon cover 

 the broken stub, leaving a blunt projection on the edge of the pinnule (Plate 

 62, Figs. 2 and 3). 



Renewal of growth has not been observed in D. ligulata, but in D. 

 media (Ag.) Grev., collected in the same region in April 1917, the first 

 cell of the assimilating hair was just protruding through the cortex at the 

 tips of many of the ultimate branches. 



At the same time that the assimilating hairs are being developed the 

 central axial row of cells of the midrib, the "original thallus" of Janczew- 

 ski (1875), is being increased in length by the addition of cells from the 

 lower side of the intercalary growing point. These cells, almost as soon 

 as formed, send out projections on either side which are cut off by cross 

 walls and develop into lateral branches. Branches of the axial row above 

 the growing point may or may not be in pairs, and do not develop from 

 every cell; but in the permanent axial row below the growing point, every 

 cell that is cut off bears two opposite branches. The axial cell elongates be- 

 low leaving the brandies on its upper part, and divides by a cross partition, 

 the branches being on the upper cell. Each of these cells elongates and 

 again divides. Thus the pair of branches will be found on the uppermost of 

 the series of four cellsi formed by two successive divisions of the original 

 cell laid down at the growing point. Soderstrom (1889) states that in 

 D. aculeata no further divisions take place in the cells of the axial row, 

 every fourth axial cell throughout the thallus bearing branches. Each 

 axial cell, therefore, must become greatly elongated to keep pace with the 

 elongation of other tissues of the thallus. In pinnules of D. ligulata, well 

 cleared and not too deeply stained, the writer has been able to trace the 

 axial row for 3 or 4 millimeters. Branches arise from every second, third 

 or fourth cell, but no wider intervals have been observed. The same holds 

 true for longitudinal sections of older parts of the thallus. Each original 

 cell does not necessarily subdivide to form four cells. For instance, in the 

 basal region of a branch of the first order, branches have been found on 

 every second axial cell, indicating that only one division has taken place. 

 In other sections, every third axial cell bears branches (Plate 59, Fig. 2), 

 inJicating that the original cell and one of the daughter cells has divided. 

 When every fourth cell bears branches, it is evident that both daughter 

 cells of the original axial cell have again divided. No greater intervals 

 between lateral branches have been found in any part of the thallus, hence 



