92 SCIENCE PROGRESS 



parietal protoplasm, which surrounds the large central vacuole. 

 Reproduction takes place by a process of vegetative budding 

 (fig. 3, e) and by the formation of numerous isogametes 

 by subdivision of the protoplasmic contents of the individual. 

 These gametes give rise to resting zygospores but are also 

 capable of germinating parthenogenetically. If the substratum 

 dries up, the protoplasmic contents of the individual subdivide 

 to form a number of multinucleate cysts, which become coloured 

 red by haematochrome and enveloped by a thick membrane. 

 We may regard this latter phenomenon as a primitive charac- 

 teristic retained from a Flagellate ancestry. It will be seen, 

 therefore, that Protosiphon furnishes a useful connecting link 

 between the more elaborate Siphoneae and the Protococcales (to 

 which Protosiphon is referred by some authorities; Oltmanns 53, 

 cf. also Wille 70a), as in both groups we have the same antago- 

 nism to vegetative division, while a form like Phyllobhun 

 of course already shows a marked siphoneous tendency. 



In recent systems of classification (Blackman and Tansley 6, 

 Oltmanns 53 and 54) it has become customary to distinguish 

 between two subdivisions of the Siphonales — the typically non- 

 septate Siphoneae and the septate (although still coenocytic) 

 Siphonocladeae. In certain forms of the latter {e.g. Valoniaceae, 

 fig- 3. f) the thallus is strictly siphoneous at first, and septa only 

 appear as the individual gets older, whilst in others {e.g. Clado- 

 phoraceae, fig. 3, g) septation is apparent from early stages 

 onwards. The general tendency is to look upon a form like 

 Valonia (fig. 3, f) or Siphonocladus (both members of the 

 Valoniaceae) as the starting-point for the Siphonocladeous 

 series ; such forms are of course readily referable to the Proto- 

 siphon-type. Oltmanns (53 and 54) has, however, propounded a 

 rather different theory as to the phylogeny of the Siphonocladeae, 

 which also influences the view of the origin of the whole group 

 of the Siphonales. This theory attempts to bring the Siphono- 

 cladeae into line with the Ulotrichales(as illustrated, for instance, 

 by the familiar genus Ulothrix) — a view which finds some support 

 in the rather close correspondence between the methods of 

 reproduction in Cladophoraceae and Ulotrichaceae. The Clado- 

 phoraceae are assumed to have arisen from a branched or 

 unbranched Ulothrix-WkQ ancestor by an increase in the size of 

 the cells, accompanied by a corresponding increase in the 

 number of the nuclei and a gradual elaboration of the chloro- 



