Howe : Phvcological studies 24!) 



plant, when they appear as a dense tomentum adhering to the 

 lower surface of all parts of the thallus except a very narrow zone 

 at the growing margin. They are commonly more or less inter- 

 tangled with filamentous, rock- boring Cyanophyceae. The 

 larger rhizoids are occasionally formed by a ventral evagination 

 of a part of the thallus-tube, the lumina of the two remaining con- 

 tinuous, but the ordinar\ T slender rhizoids, the measurements of 

 which are given above, appear from an early stage to be external 

 appendages to the thallus-tubes, their basal cells standing in about 

 the same relation to the walls of the main cells as do the small 

 fibular cells of Siphonocladus Hgidus shown in plate 14, fig. 4. 

 Like Siphonocladus rigidus, Petrosiphon adhaerens is accom- 

 panied by a fungus, or, at least, a chlorophylless filamentous thallo- 

 phyte, though this is not so uniformly present as in that species ; 

 we have never dissected a thallus in which it could not be found, 

 yet large parts are sometimes destitute of it and we are of the opin- 

 ion that its relation to the alga is that of a parasite rather than that 

 of a subordinated symbiont. The external hyphae lie somewhat 

 loosely among the thallus-tubes or rhizoids ; they are fuscous, 

 septate, 2.5-5// broad (cells mostly 2-15 times as long), for the 

 most part sparingly branched, straight or sometimes contorted or 

 torulose ; the hyphae finally penetrate the walls of the thallus-tubes 

 and follow their cavities in the direction of their growth, mean- 

 while becoming lighter-colored and forming much-elongated, 

 ribbon-shaped or virgate-fasciculate, decompound clusters of 

 branches ; the cells of these endophytic branches are mostly 5—1 2 u. 

 long, ellipsoid, oblong, or ovoid, and are occupied chiefly by a 

 large vacuole, with 1-3 small refringent bodies usually lying close 

 to the wall. Nothing which could be identified with reasonable 

 certainty as a reproductive body of this fungus has been observed. 

 The only described species, so far as we can discover, which 

 may suggest itself as possibly congeneric with Petrosiplwn adhaer- 

 ens is the minute SipJwnocladus vohiticola Hariot, which grows on 

 the shells of Voluta in Tierra del Fuego. But this, according to 

 the original description and figures, * and according to supple- 

 mentary descriptions and figures by Bornet, f is quite different and 



* Jour, de Bot. i : 56. 1887 ; Mission scientifique du Cap Horn 5 : 22. pi. 1. f. 

 2-4. 1889. 



tBull. Soc. Bot. France 36: clix, clx. pi, 10. f. /, 2. 1SS9. 



