* 
60 pees 
and generally soon fall off; more rarely they persist also in the older 
portions, Every joint bears such a hair at the upper transverse wall, 
at the margin between two pericentral cells. Therefore these also 
stand in regular t-arrangement. In falling off they each leave behind 
mber at irregular intervals, so t long internodes of 18 
more cells alternately occur. e branches may repeat 
branched. A distinct main axis is not present. ‘The cell- alls ate 
somewhat inclined to swell. 
The following short table gives information with regard to the 
length and diameter of the joints :— 
Length. Diameter. 
B B 
Apical cell ... sal i ae an 9 
6 joint from the top a ee ae 12 
10 ” ” ”? $e4 5 eis 15 
15 ” ” ” Nee ena 23 
 . S a ae a 34 
From this point the joints increase in length and diameter 4 
very slowly, as may be observed from the following table of leng™® | 
and diameters :— 
Distance from Apex. Length. Diameter. 
m. BB 
0°08 - 8 20 
0°16 : 12 até 21 
0°24 ws 16 
0-41 ‘ia 40 53 
re 2 62 Ne 
7. 92 
2°5 . 105 a 
5°0 a 160 a 
8:2 180 Me 
12-0 ‘ 188 bs 
It is seen that here the diameter of the joint is ab mit almost 
than the length. Further on, the growth of the diametef joints 
the growth of the diameter may also continue longer, a8 1m Bi A 
mentioned case of “The Algv of the ‘Gazelle,’” and the older joints 
then as long as thick, * 
_ The branches bearing tetraspores (Plate XVI., Figs. 5, oni 
slightly altered in comparison with the normal, sometimes _ 
bent to and fro, usually almost straight. ; io of the 
The antheridia (Plate XVII., Fig. 1) are seated a8 9 Mee of the 
second cell of a hair, so that their stalk is bicellular, con®) in It 
short basal-cell, and the first somewhat elongated cell © © pect 
one instance I found on a hair two antheridia—viZ., 1%. shove! 
represented in Plate XVIL., Fig. 1. The portion of the rs vi 
antheridium easily falls off, but the antheridium itself 18 °° 
