18 Prof. T. G. Bonney — The TwtSill Conglomerate. 



that the minute ramifying tubes which are found throughout tlie Loess 

 were the relics of roots of grasses, but this is a mere hypothesis, 

 and one which is by no means probable. That a succession of 

 plants growing in one area for ages should have left no debris but 

 these fine tubes, and that they should occur throughout great thick- 

 nesses of Loess, where it is 500 and 600 feet in depth, seems quite 

 incredible ; and I am more than inclined to accept Mr. Geikie's 

 tentative opinion when he says, " I am not aware that any trace of 

 vegetable matter has ever been found in the tubes, and the capillary 

 structure, like the concretions, may be of inorganic origin. Chemical 

 analyses, at all events, have shown that Loess contains little or no 

 organic matter, which we might have expected to meet with in much 

 greater abundance had plants given origin to the innumerable vertical 

 pores which, are so commonly present in the typical deposit of the 

 Ehine and the Danube" (Prehistoric Europe, p. 237). Again, the 

 fact that the Loess is deposited with singularly uniform composition 

 and structure, independently of the nature of the subjacent rocks, 

 shows that if in situ and not largely transported, it is entirely 

 different from other subaerial soils which partake of the local charac- 

 teristics of the rocks on which they rest. The fact is, the more we 

 examine this solution, the more impracticable it becomes. By a 

 process of exhaustion we have pretty nearly sifted every possible 

 source for the Loess. 



{To be concluded in otcr next Number.) 



Errata. — Geol. Mag. Dec. II. Vol. VIII. p. 569, last line, for " merely " read 

 " surely " ; p. 572, line 10, for "with honoiuing " read " in honouring." 



IV. — On the Twt Hill Conglomerate. 

 By Prof. T. G. Bonney, M.A., F.E.S., P.G.S. 



INCE the date of my communication to the Geological Society on 

 the sections at and in the neighbourhood of Twt Hill,^ I have 

 more than once visited the district, and have spent much time in 

 working over the specimens which I have collected. The exact 

 relation of these conglomerates to the granitoid rock on which they 

 rest is a matter of great importance as bearing on theoretical 

 questions relating to metamorphic changes in stratified rocks — 

 questions at which for the last three or four years I have lost no 

 opportunity of working. I should therefore, in all probability, have 

 recurred to the subject, even if my attention had not been directed 

 to it by my friend Dr. Eoberts' communication and criticism of the 

 above paper.^ In the latter, it may be remembered that we assigned 

 the Twt Hill conglomerate to the granitoid series. The reasons 

 which led myself, at any rate, after much consideration, to the con- 

 clusion, were chiefly the following : — 



1. That after very careful examination in the field I was unable 



1 "Written conjointly with Mr. Houghton, Quart. Journ. Geol. Soc. vol. xxiv. p. 321. 



2 See Geol. Mag. Vol. VIII. Dec. II. 



