80 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
actually attaining that of spherical separation, thus producing the 
bales shown in the Rio Negro section, characteristic of many weather- 
ing basalts, an effect coupled with the want of definite stratification 
within the stratum. 
Once the tillite beds underlying sandstones are reached by dissecting 
streams or rivers, the overlying strata widen out with steep cliffs, the 
small streams falling over these rocks have falls, and, owing to the 
ready removal of the tillite and the growth of the main valley in this 
section, the side streams present the appearance of hung up valleys. 
Age of the Boulder-Beds.— As commonly stated on the authority of 
Dr. Derby the boulder-bearing deposits are to be regarded as the 
equivalent of the Upper Carbonic or Permian glacial period now 
recognized in India, Australia, and South Africa. This Permian age 
of the beds in Brazil was accepted by Professor Branner in his Geologia 
elementar (Rio de Janeiro and Sao Paulo, 1906), and still more 
recently by Dr. I. C. White in his Report on the Brazilian coal field. 
In the same monograph, Mr. David White presents the most complete 
description and analysis of the flora of the south Brazilian field yet 
published. From this report it appears that the flora succeeding the 
conglomerates on the south in Santa Catharina are of the Lower 
Gondwana type which immediately succeeds the glacial beds in India 
and elsewhere in the eastern hemisphere. Dr. David White corre- 
lates the Jaguaricatu beds [Orleans conglomerate] and associated 
sandstones and shales of Dr. I. C. White with the Talchir beds of 
India, the Dwyka conglomerate of South Africa, and the basal 
conglomerate of the Permian series of Argentina. 
Divisibility of the Glacial Beds—— The question is naturally asked 
whether there was one or more than one episode of glaciation or ice- 
action in southern Brazil. The facts concerning the distribution of 
conglomerates in this field are as yet too little known to enable one to 
give a satisfactory statement in reply to such a question. The 
localization of the typical tillite beds near the base of the series so 
suggestive of local contributions of débris complicates the question. 
If the conglomerate beneath the whitish sandstone at Serrinha Station 
with its glaciated fragment be of Permian age and pass, as it appears to, 
beneath the boulder-bearing beds exposed westward and southwards 
towards Lapa, then in that section there are evidences of two epochs 
of ice-action separated by the deposition of the sandstone beds which 
form the walls of the gorge of the Iguasst. 
The tendency of the boulders to display themselves in a marked 
manner in certain zones often only a few feet apart is evidently a 
