478 . The American Naturalist. [May, 
the lakelet would become much more permanent, and its basin would 
be apparently rock-bound, with its surplus water flowing over a rocky 
outlet. ; 
Mr. Marr concludes since many of the Tarns he examined are in- 
stances of the third class described above, that the lakes of that region, 
at least, give no support to the theory that the basins in which they 
occur were hollowed out by ice. (Quart. Journ. Geol. Soc. Feb., 
1895.) 
The Loess of Northern China, —The superficial deposits of 
Shantung formed the subject of a paper by Messrs. Skertchly and T. 
W. Kingsmill read before the Geological Society of London at a recent 
meeting in which some interesting facts were made known concerning 
the Loess of that region. The Loess east of the Pamirs is extensively 
developed over an area of over one million square miles. It is some- 
times over 2000 feet thick, and occurs up to several thousand feet 
above sea-level. Evidence was brought forward by the authors with 
the intention of establishing the absolute want of connection between 
the Chinese Loess and the present river-systems, its original stratified 
condition (as shown by variation of tint and horizontality of layers of 
concretions) and its subsequent rearrangement to a great extent. The 
absence of marine shells was discussed, and the suggestion thrown out 
that the shells had been destroyed by percolating water. The authors 
gave their reasons for supposing that the Loess is a marine formation, 
and stated that the sea need not have reached to a higher level than 
600 feet above the present sea-level, for the Pamir region where it 
occurs, 7000 feet above the sea, is an area of special uplift. They 
maintained that there are no proofs of the glaciation of Northern and 
Eastern Asia, so that Chinese Loess could have no connection with an 
area of glaciation. The zoological, ethnological, historical and tradi- 
tionary evidence alike pointed to the former depression of Asia beneath 
the sea, and the subsequent dessication of the land ee upon 
re-elevation. (Nature, March, 1895.) 
Geological News, Pa.rozorc.—In studying the remains of 
Radiolarians and Sponges in the precambrian rocks of Bretagne, M 
L. Cayeux arrives at the following conclusions : 
(1). There exists at the base of the pea of Bretagne numer- 
ous sponge spicules representing many speci 
(2 )- All, or at least nearly all, the oe of the siliceous sponges 
were in existence at this early period. 
