JIOXOGRAPn OF THE LABOULIJEMACKJ:. 231 
the wall-cells, although qnite imlopeiident of it, and the same Is in general true of all 
the genera. In forms in whicli, like Ceratoinvces, the wall-cells become very numer- 
ous, a corresponding increase in the number of the canal-ccll.s (akos 2)lace ; allhough 
the number formed may be fewer, as is indicated in Plate XXIV, fig. 8, which rc'2)re- 
sents the growing apex of a perithecium in section, the divisions of the wall-cells (frc) 
being evidently more numerous than those of the canal-cells (c c). There seem, also, 
to be certain variations in the relative position of the canal-cells. In Sphaleromyces, 
for example (Plate III, fig. 2), the cells (.r) are probably the lower canal-cells which 
have, by pushing between the wall-cells, assumed an apparently external position ; and 
some similar modificaiion of the more normal course of development may account fur 
the peculiar arrangement of the cells at the tip of the perithecium in this genus, when 
seen antero-posteriorly as in fig. 1. It is not improbable that there are, in certain 
cases, variations of the course of development above described : yet in all the instances 
in which a careful examination has been made, it seems to correspond in all essentials. 
"While the waH-cells of the perithecium have a definite protective function, persist- 
ing and forming an envelope around the ascus and spore masses, and regulating and 
directing the disj^ersion of the spores; the inner series, as we have seen, performs an 
entirely different function which may well be compared, in so far as concerns the cells 
which I have called the parietal cells, to that of the so-called tapetal cells in the sporangia 
of the higher cryptogams ; while the cells which 1 have called the canal-cells bear a 
similar resemblance in function to the neck-cells, for example, of an archegonium. It 
is needless to say, however, that there is not the slightest homology between the two 
structures in either case. The function then of the parietal cells is to make room for 
the developing ascus masses, and allow them to float free within the cavity of the 
perithecium ; while that of the canal-cells is to afford a channel of exit by means of 
which the spores may make their way out through the pore of the perithecium. For, 
as we have seen in the case of Stigmatomyces, which is typical of the family generally, 
the growing ascus masses press upon and destroy not only their own connections with 
other cells above and below, but also the thin-walled parietal cells around them; while 
the mass of mature spores, which is constantly being pushed upward, has a similar 
effect upon the canal-cells, the latter, as a rule, having disappeared entirely when the 
discharg 
In exceptional cases, as in that of Stigmatomj 
w\alls of certain of the canal-cells (Plate I, fig. 24, hc) appear to become somewhat 
indurated, and to regulate, to some extent, the spore discharge ; but I have observed 
no other instance than the one just mentioned, unless it be in the ca^e of Sphalero- 
myces (Plate III, fig. 1) above alluded to. 
