107 
aerial mycelium of some races produced clumps (PI. XIII, XXII, XXIII), 
which under the microscope are seen to be due to a peculiar distortion and 
abundance of the aerial mycelial tips (Fig. 6, d). This peculiar behavior 
of the terminal parts of the mycelium shows some similarity to the branch- 
ing figured by Ward (123) in Botrytis in the early stages of development of 
attachment organs. Anastomosis is very common with this fungus (Fig. 
5, a and b), and Ward (123, fig. 19) has figured anastomosis of very similar 
character for Botrytis. Many citations of its occurrence are given by 
Beauverie and Guilliermond (13). (See also their figures 4 and 8.) 
The nuclei in the mycelium are extremely small, but may be seen readily 
when stained with gentian violet, and still better if stained with iron-hae- 
matoxylin. They vary in number, but are never less than two and usually 
more; they do not typically group in pairs; and they are irregularly dis- 
tributed in the protoplast. Nuclei apparently in mitosis are frequently 
seen, but since they are so small no details were noted except that the 
mitoses of all of the nuclei in one cell seem to be simultaneous, and mitosis 
was probable in adjacent cells. Reports on the nuclear conditions in the 
fungi imperfecti are few and unsatisfactory, doubtless owing to the extreme 
difficulty of the subject. Dangeard (39) notes nuclei and mitosis in the 
rather anomalous genus Bactridium; Beauverie and Guilliermond (13) in 
Botrytis; while Higgins (66) gives quite satisfactory figures for Mycosphae- 
rella. 
Senescence phenomena of aerial mycelium (Fig. 7, a—b).—When the 
aerial mycelium is young it constitutes a more or less abundant, loose, 
arachnoid, fluffy mass. In quite old cultures it is observed to mat down 
close to the surface of the medium in a thin, glazed, dead layer. Inter- 
mediate between these two extreme conditions interesting phenomena 
occur. The first observable change from that of the normal, vigorous 
mycelium is that certain cells of a filament, often many adjacent cells, be- 
come nearly or quite devoid of protoplasm (Fig. 7,7, 0, and g), though cells 
at each end of such a series still appear normal (see 7 and j). Quite fre- 
quently the fungus re-grows from a protoplasmic cell, through the empty 
threads, as is shown in » and 0. In other instances, and much more com- 
monly, the empty cells gradually collapse until they remain as very 
thin, smooth filaments (¢, 4, and m), apparently of gelatinous texture. 
Where two filaments undergoing such dissolution cross they blend (a, 0, 
mu); and where several meet, rather large amorphous unorganized masses 
are seen, superficially much resembling a plasmodial structure (a). There 
was, indeed, at first, suspicion that there might be present a plasmodial 
