E. H. Eoivorth—The Loess. 347 



the accumulations of vast depths of Loess through which the hones 

 are distributed irregularly, and I must repeat that I cannot conceive 

 how the wind could transport them or the dust bury them intact, 

 unless the meteorological conditions were entirely different from 

 those now in existence, and quite transcendental. 



Let us now turn to the structure of the Loess. Baron Kichthofen 

 accepts the theory which makes the ramifying calcareous tubes 

 found in the Loess to be the root impressions of grasses, and bases 

 a good deal of deduction upon it. The theory that they are so, and 

 are not due to calcareous filtration, is at present, however, very 

 much in need of evidence, and, to my mind, has utterly collapsed. 

 I would add to what I said on this subject in a previous paper, that 

 if these tubes were mere casts of roots, we ought assuredly, in regard 

 to & forest flora, to find any number of casts of tree roots, and roots 

 of herbaceous plants, which are so easily discriminated ; but we find 

 none of these — only the ramifying fine tubes which have been called 

 grass roots, but which, to my mind, have nothing whatever to do 

 with grass roots, but are due to the percolation of water charged 

 with calcareous matter. This is further supported by the fact that 

 they do not occur where the Loess has a dense hard structure — only 

 where it is porous ; that they occur chiefly near the surface, and 

 diminish in size as we go down, which is the exact result we should 

 expect from percolating threads of water acting on a deposit highly 

 charged with carbonate of lime, and inconsistent with a deposit 

 gradually growing higher, while the plants, whose roots the tubes are 

 supposed to be, were continuously being buried by additions of wind- 

 borne dust. In regard to the features just named, Mr. Todd 

 (who, by the way, does not question their being root-marks) 

 says : " They may be said to vary inversely as their distance below 

 the surface. Near the surface, besides being most abundant, larger 

 ones are found. At the depth of 30 or 40 feet, they are very minute 

 and rare. None have been found lower than about 45 feet, although 

 several favourable localities have been examined " (Todd, op. cit. p. 8). 



The concretions that occur in the Loess add very considerably to 

 the weight of evidence which makes these tubes to be the results 

 merely of percolating water, for these concretions are clearly the 

 result of the same agency acting slightly differently. In regard to 

 these concretions, I shall be pardoned for quoting a graphic description 

 by my correspondent. Prof. Ellsworth Call. He says: "They assume 

 all possible shapes from the spherical through the spheroidal to the 

 oblong ; in all cases they are more or less numerously studded with 

 roughened projections. No one shape seems to obtain more than 

 another, and not unfrequently several are found cemented together, 

 forming an eccentric single mass. They are certainly characteristic 

 of the Loess, for that formation nowhere occurs without their pi'esence. 

 They are decidedly hydraidic, as loould be naturally inferred from 

 their constitution. In no case have I ever observed fossils — either 

 moUusks or vegetable matters — acting as a nucleus. On one 

 occasion 2803 of these bodies were crushed with that especial point 

 in view. In nearly every instance, 2789, they were found to contain 



