An Eocene Microthyriaceous Fungus from Mull, Scotland. 69 
the Microthyriaceae, the Hemisphaeriaceae and the Trichopelta- 
ceae in a new order, the Hemisphaeriales, which are all super- 
ficial, and are thus sharply marked off from the Dothideales, 
in which the ascostromata are developed beneath the cuticle. 
For the hemisphaerical fruiting bodies he adopts von Hohnel’s 
term thyriothecium. The distinctive characters of the Micro- 
thyriaceae, with which we are particularly concerned here, are 
the radiate (very rarely linear) dimidiate thyriothecia, on the 
“inverse”” method of whose development von Hodhnel and 
Theissen lay great stress. They regard the thyriothecium as 
corresponding to the basal half of the spherical body such as 
occurs in related families (e.g. Perisporiaceae) which has become 
inverted against the leaf surface. The family is subdivided 
according to the presence or absence of a free mycelium, the 
number of cells and colour of the ascospores, the presence or 
absence of paraphyses, and so on. This classification is adopted 
with slight modifications by E. M. Doidge (1920) where a 
résumé of the general characteristics of the group will be found. 
Arnaud (1918) proposes a somewhat different classification 
based on that of Spegazzini, and uses the term Microthyriaceae 
in a rather wide sense, while excluding, however, certain genera 
such as Amazonia. Arnaud gives the pycnidial forms a special 
name (Microthyriopsidaceae) ; the pycnidia resemble the thyrio- 
thecia as a general rule, and therefore we cannot be quite sure 
which we are dealing with in the present case. Certain spore-like 
bodies on the leaf cuticle might perhaps be conidia. 
The form of the thyriothecia in the fossil is very similar to 
that of Microthyrium itself, while in this and allied genera the 
mycelium is absent or evanescent. Presuming the ascospore to 
belong to the same species as the fruiting bodies, a relationship 
is indicated to the genus Phragmothyrium of von Hohnel (1912, 
P. 347) which was instituted to include “ Microthyrium species 
with more than 2-celled spores.”” Amazonia Theiss, is also phrag- 
mosporic (5-celled) but has a persistent mycelium and rather 
different thyriothecia with fimbriate margins. It is therefore 
suggested that the name Phragmothyrites be used for fossil forms 
belonging to the Microthyriaceae, the exact position of which 
is uncertain, but which appear to be most closely related to 
Phragmothyrium as defined by von Héhnel. The Mull specimen 
may be described as Phragmothyrites eocaenica n.sp. Thyrio- 
thecia circular, radiate, with entire margin, scattered singly 
(rarely concrescent) on the leaf surface; stigmocysts circular 
and deeply crenulate ; ascospores (?) 5-celled; mycelium ? absent 
orevanescent. Occurring on coniferous leaves (Pityophyllum sp.) 
in beds of Lower Eocene age, Mull, Scotland. 
I have only succeeded in finding a few other references to 
