306 DIVISION II.—COURSE OF DEVELOPMENT OF FUNGI. 
have another form of basidium, as was first observed by Tulasne. Subhymenial 
hyphal branches swell in these genera to a spherical or ellipsoid cell filled with 
protoplasm, the initial or primary basidium. This cell then divides by vertical 
longitudinal walls usually into four cells arranged as the quadrants of a sphere, 
secondary or definitive basidia, each of which then puts out a long sterigma and abjoints 
a spore upon it which absorbs the whole of its protoplasm. Slight variations occur 
in the number of the secondary basidia formed from a primary. But more important 
than these are the differences in the separation of the basidia, which may go so far 
that before the spore-formation each of the three sister-basidia becomes separated 
down to the base, or, on the other hand, all may remain united as at first; or again 
the division of the primary basidium may be incomplete or be omitted, and the 
basidium become imperfectly chambered or lobed, each division answering to a 
sterigma ; this is the case, according to Tulasne’s account, in Tremella violacea and 
T.Cerasi and in Sebacina incrustans, and according to Brefeld in Tremella violacea !. 
Cases of the latter kind form the transition to the two-lobed or four-lobed basidia of 
Dacryomyces and to ordinary Hymenomycetes, 
In Auricularia Auricula Judae (A. sambucina, M.) the primary basidium is long 
and cylindrical and very like the basidia of Dacryomyces and Calocera. It divides 
by transverse walls into a row of four or five daughter-cells, each of which sends out 
an erect subulate sterigma, which rises above the surface of the hymenium and 
abjoints a spore. The sterigma issues from the apex in the uppermost basidium of a 
row, in the rest from the side close beneath the upper wall; the formation of 
sterigmata and the abjunction of spores begins in the uppermost basidium of a row; 
the rest follow it in order from above downwards. Exactly similar phenomena are 
described by Tulasne in Hypochnus purpureus, only the end of the row is in this case 
hooked and the terminal cell itself is sterile. The shape of the ripe spore also is pecu- 
liar, being renzform in most of the ‘Tremellineae. These characters necessitate the 
separation of the Tremellineae as a special division of the Hymenomycetes. The 
gelatinous constitution of the sporophore is a convenient character in many cases, 
but would not be of sufficient importance by itself, the more so as Sebacina incrustans 
and Hypochnus purpureus, the latter of which I have not myself examined, do not 
appear to have gelatinous membranes, 
By far the greater number of the Hymenomycetes form only one hymenial layer 
on each sporophore, whether it is of only brief duration or lasts longer and even for 
many years. The course of its development, the maturation of its parts, exhibit 
in general the same progressive advance towards the margin and apex as has 
been already described in connection with the growth of the whole apparatus which 
bears the hymenium. A few slight deviations however occur. On the one hand, 
the definitive completion and differentiation of the originally uniform elements of 
the hymenium are effected, according to Brefeld, in Coprinus simultaneously at all 
points of the hymenial surface; it has long been known that in C. micaceus and 
C. comatus the ripening of the spores, as shown by the lamellae turning black, even 
begins at the margin of the pileus and the edges of the lamellae and advances to the 
middle of the pileus and the base of the lamellae. On the other hand, indubitable 

1 See particularly the Ann. d. sc, nat. XV (1872), p. 234. 
