320 



and, as the ends of each of the branches are placed distichously alter- 

 nately on both sides of the main axis, the whole sympodium gets a 

 very regular, feather-like appearance. An exception from this rami- 

 fication is only made by the second side-branch (the first of the three) 

 this being firstly branched once, the two branches originating from 

 this branching are then sympodially ramified in the way described 

 above. FALKENBERG gives very fine illustrations (1. c., pi. 17 figs. 13, 

 14) of these lateral sympodia. The main axes of these four sympodially 

 ramified branch systems become polysiphonous with the exception 

 of their uppermost ends. From the monosiphonous side-branches of 

 these lateral sympodia the net is developed. This is formed in such 

 a manner that outgrowths emerge from two neighbouring filaments 

 and grow together after which the outgrowths are separated by walls 

 from the mother-cells. But this growing-together process is not re- 

 stricted to filaments from the same group of sympodia, for the peri- 

 pheral filaments of one group of sympodia connect themselves with 

 those of other groups (cp. Fig. 331 a). 



Fig. 330 a shows a transverse section of the thallus of Dicty- 

 ums occidentalis. This is seen to be quadrangular with concave sides ; 

 but, because of the spiral arrangement of the net, this is cut through 

 and wanting at the one side, and further, because of the very thick 

 cortical layer of the main stem, the base of the two branch systems, 

 drawn in the figure, is not clearly seen. When FALKENBERG, regarding 

 Dictyums occidentalis, says, I.e., p. 681, Mass an dieser iiberhaupt 

 schwacheren und schlankeren Dictyurus-Art die Hauptaxen der Sei- 

 tensympodien an der sterilen Pflanze wenigstens durchweg 

 monosiphon bleiben" it does not correspond with my observation (cp. 

 Fig. 3306) the main axes in my specimens being polysiphonous 

 with four pericentral cells. It is only in small and very feebly developed 

 branches that I have found the side-branches throughout monosi- 

 phonous. And besides it is so in the case of the original specimens 

 from Mexico, at any rate in that specimen I have examined. Of the 

 four main axes in each branch-system the two run out in the pro- 

 longed corner of the net, the other two end in about the middle of 

 the concave side (Fig. 330 a). And here Dictyums occidentalis shows 

 a marked difference from Dictyums purpurascens according to FAL- 

 KENBERG'S description. In the latter we have between the 

 larger acute corners smaller obtuse edges to which the polysiphonous 

 axes run out. This I have not found in my material ; between the high- 



