116 
are parallel to its long axis and which, diverging on the surface, clothe 
it with a kind of felt. The livphse on the inside which line the neck 
canal also bend outward and form hairs. These are all directed ob¬ 
liquely upward so as to permit the escape of the spores but to guard 
the entrance to the perithecium. In the older as well as the earliest 
stages the inner wall is composed of a colorless thin-walled pseudo- 
parenchyma, and surrounds the entire cavity of the perithecium with 
the exception of the neck. The origin of this cavity and of the neck 
canal can not be definitely decided. The paraphyses and asci arise 
from the inner wall; the former are present in large numbers and are 
very long, slender, and thin walled. Their contents consist of finely 
granular protoplasm containing many globules that reflect the light. 
In their earlier stages the asci are composed of extremely delicate 
club-shaped cells filled with strongly refracting protoplasm. Their 
wall remains very delicate even when they are fully grown. The form 
of the mature ascus is that of a cylinder, obtuse at the upper end and 
suddenly contracted below into a short, slender pedicel. It contains 
eight spores, which are oval, dark-brown, and three-septate at maturity. 
The two middle cells are each again divided by an oblique wall, which 
disappears when the spore is turned through an angle of 90° (Fig. 6)* 
Very rarely there are a greater or less number of cell walls than stated 
above. The spores are slightly constricted in the plaues of the septa, 
and the whole is surrounded by a thin, gelatinous layer. The length of 
the spore averages 14-18//, the diameter 5-9//. 
The details of the perithecium having been fully examined, the point 
has been reached when the systematic position of the fungus should be 
more thoroughly considered. The main points upon which this depends 
are the following: The brown spores with their longitudinal and trans¬ 
verse septa; the presence of paraphyses; the existence of a stroma, if 
only an imperfectly developed one, which is certainly Valsoid, as is 
proved by the fact that it is very small in circumference and has peri- 
thecia projecting from the center and pyenidia from the circumference 
(Fig. 7). Of all known genera Fenestella is the only one which can be 
considered, since it is characterized by the above-mentioned points. 
Besides, there is no reason why the fungus should not be placed in this 
genus. It should be mentioned, however, that in most species of 
Fenestella . especially the native species F. princeps , the spores have many 
more septa; yet Saccardo mentions those having but few. This species 
differs from all other Fenestellw in the great development of the neck, 
which gives it the appearance of a Valsa, and when the color of the 
spores is taken into consideration it stands very close to a Pseudovalsa . 
It may therefore be decided, merely on account of the longitudinal 
septa (which, as stated above, are visible only in a certain position of 
the spore) that it should be classified not as a Pseudovalsa but a Fenes¬ 
tella. Although the presence or absence of two walls is a very small 
matter in itself, and certainly one to which no great systematic value can 
