328 BOTANICAL GAZETTE [NOVEMBER 
knob-like or filamentous processes in lieu of holdfasts on clean spots 
on the gelatine (fig. 8), whereas those in contact with particles of dirt 
were more or less branched (jigs. 9, 10). On roughened glass many 
zoospores developed disks (fig. 12). On clean glass the regularity 
of these disks was very striking (fig. 11). Irregularity of surface 
evidently induces irregularity of growth and form. This appears 
even more plainly from the experiments on marine algae (p. 343). 
Za 
Fics, 1-12.—Germinating zoospores of Oedogonium * ? X780. 
1, 2, 3, Germinated on the surface of water (1, 2 filtered, 3 unfiltered) showin: g extremely ee 
holdfast.—4. Same somewhat older, ater into two cells.—s, 6. Same but with hypha or 
' outgrowths at the basal end of each spo Bacteria surround these filamentous eae —o. Ger - 
on freshly split mica. Note well- eviaeet holdfast.—8. Germinated on clean wet pase aa) 
minated on a spot of dirt on wet sa —1o. Germinated on spots of dirt G inated eo 
tine.—11. Germinated on clean cover glass ote extremely regular holdfast—12- i d 
side of roughened glass dish. Note ae holdfast. 
Even clean glass and freshly split mica are not absolutely smov® 
The dustless upper surface of clean water is smooth. . Pa : 
wet gelatine, as Prrerer’s work on tendrils showed (16). On 4 
two no holdfasts are formed by germinating zoospores. 
holdfasts longer or shorter hypha-like filaments appeats 
short indeed, and branched only where the surface is ro holds 
dust or dirt. On carefully polished clean cover glasses, h 
of remarkably regular form were developed. As 
later (p. 344), uniformly ground cover glasses used in expe 
often VOY 
y 
