410 
Proceedings of the Royal Society 
by the absence of an articulated stem or its representative the 
centro-dorsal plate ; by its viscera being lodged in a hollow 
peduncle with a continuously calcified wall ; and by the absence 
of an anal opening. 
In 1846 Professor Steenstrup described under the name of Gya- 
thidium, a genus of fossil Crinoid from the chalk of Faxoe, which 
occupies a debatable position between the base of the Eocene Ter- 
tiaries and the top of the Cretaceous series. The only portion yet 
described of Gyathidiuru is a deep cup or tube with a spreading 
base of attachment and a thickened rim with articulating facets 
for five arms. The cup of Cyathidium is somewhat more sym- 
metrical and coralloid than that of the recent West Indian form ; 
but I see no distinction between them of generic value, and I think 
we must accept Holopus as another of the links which recent in- 
vestigations have made so numerous between the faunae of later 
geological periods and that of the present time. 
2. On the Diurnal Oscillations of the Barometer. — Part II. 
By Alex. Buchan, M.A. 
In this communication the author stated that he limited his 
remarks on the present occasion to some of the more prominent 
results he has arrived at in the course of this investigation. The 
paper itself, with the tables, will be submitted when the computa- 
tions have been finished and thoroughly revised, — a work which 
must necessarily yet take some considerable time. 
It is proposed that Part II. consist chiefly of tables showing the 
arithmetic mean values of the hourly variations of the different 
months of the year, with remarks on the more evident conclusions 
which may be drawn from them. The number of places for which 
data of more or less completeness have now been obtained exceed 
130, situated in different parts of the globe. To these it is proposed 
to add the results of observations made at sea, chiefly those made 
by the “ Challenger ” and “ Novara ” expeditions. 
As regards temperate regions, such as Great Britain, periods of 
no more than three years’ observations give only the broadest 
characteristics of the diurnal barometric curve. From 20 to 
25 years will probably be found to be required to show the 
i 1 
i 
