July is, 1914 
Fusarium on Sweet Potato 
273 
because their delicate membrane can not withstand the increasing 
pressure to the same extent. 
The influence of water has to be fully Studied in order to understand 
these morphologic changes. The peridium of perithecia may be almost 
smooth when the lack of water prevents further growth of its cells. If 
the water and food supply, however, allow further development, pro¬ 
tuberance-like projections will show the permanent activity of the peri- 
pherical cell complications. Also the periphyses will prolongate and 
form a compact cone adorning the perithecium with a beak. This beak 
is almost colorless and contrasts therefore with the red color of the perid¬ 
ium (PI. XV, fig. A), The ostiolum at the top of the beak accumulates 
the ascospores as soon as the gradual drying out of the medium exerts a 
pressure on the perithecium. The peridial cells dry out and shrink, 
pressing the content of the ascus ball through the throat out of the ostiolum. 
Here they accumulate in a brown mass (PI. XV, fig. A , 2 and 4). If the 
ostiolum is closed so completely that a higher pressure is required, the 
ejaculation is more explosive and the ascospores are shot out to several 
millimeters distance. This agrees exactly with the description given by 
Dr. Erwin F. Smith for Neocosmospora. Also, the stroma of our 
fungus is as variable in every respect as in Neocosmospora. It may be 
reduced to a minimum or grown as a thallus-like effuse layer. 
Hypomyces ipomoeae may be a good example for discussing taxonomic 
difficulties. The diagnosis of this sweet-potato fungus, based on pure 
cultures, agreed closely with that of Nectria cofjeicola Zimmermann (1901), 
a saprophyte on Coffea and Theobroma, also with Nectria Goroschankini- 
ana Wahrlich (1886) from roots of Vanda tricolor , to some extent 
with Nectria cancri Rutgers (1913) from Theobroma. All these fungi 
were well illustrated by the authors, but mostly from field material. 
The differences shown in the following list between measurements from 
different authors will be understood from a discussion of their origin. 
There are two methods of measuring spores, one of which gives the 
absolute size, the other the average size; the former comprises a much 
larger fluctuation of the spore size than the latter. The absolute fluc¬ 
tuation, therefore, is likely to include young and immature spores, which 
are smaller and less characteristic than normal mature spores. This is 
especially the case when the relations between size and age and septa- 
tion are neglected. Immature ascospores may be comparatively small, 
3 to 5/i in diameter, but broader, 5 to 6/z, in maturity, and still broader, 
6 to 8 pi, in germination or from overwatering. The absolute fluctuation, 
3 to 8 /a, is too large to be of any taxonomic value, but the fluctuation of 
the average breadth, 4.5 to 6/*, has some value because it reduces in a 
measurable way the limits of the absolute fluctuation. This reduction 
can be performed by measuring repeatedly hundreds of spores and taking 
the average size of each hundred. With cultures in different ages, 
