856 



SCIENCE. 



[N. S. Vol. XIX. No. 492. 



the Portland cement industries these deposits, 

 which up to that time were u^ed only as a 

 source of land plaster, have become of great 

 importance as a source of plaster of Paris; 

 and, although it is not as pure as Nova 

 Scotia gypsum, it finds a ready market as a 

 wall finish. Plaster of Paris is used at pres- 

 ent in place of lime plaster as a wall finish on 

 account of its quick setting, so that the build- 

 ings may be occupied without delay. It is 

 also extensively used in the manufacture of 

 Portland cement to retard the set, and, if not 

 more than two per cent is used, it adds ma- 

 terially to the final strength. The manu- 

 facture of plaster of Paris depends upon 

 driving off part of the water from gypsum, 

 care being taken not to expel all the water, 

 in which case the plaster fails to set. The 

 hydrate formed in this way is known as plaster 

 of Paris and has the property of again com- 

 bining with as much water as has been driven 

 off and forming a hard network of fine crys- 

 tals, or, as it is called technically, the property 

 of setting. The present processes of manu- 

 facture were then described, and an historical 

 sketch of primitive and former methods was 

 given. 



Mr. Wm. A. Bryan, ornithologist of the 

 Bishop Museum, Honolulu, described the work 

 he was doing in cataloguing and describing 

 the academy's Menage collection of birds of 

 the Philippine Islands. This collection of 

 over 4,500 birds was made by D. C. Worcester 

 and F. S. Bourns in 1890-93, and was the 

 best source of material for Mr. Bryan's pur- 

 pose of working up all the birds of Polynesia 

 as he has already done for the Hawaiian 

 Islands. H. Gale, 



Secretary. 



THE GEOLOGICAL SOCIETY OF WASHINGTON. 



At the 156th meeting, held April 27, 1904, 

 Mr. N. H. Darton presented a paper on ' Salt 

 Lake South of Zuni, New Mexico,' illustrated 

 by stereopticon. Mr. Darton described the 

 topography and geology of this interesting 

 lake basin, and presented various theories for 

 its origin. His paper will shortly be pub- 

 lished in full. 



Mr. George H. Ashley then discussed the 



plain around Middlesboro, Ky., and its rela- 

 tion to the Appalachian structure of the re- 

 gion. A study of the general structure shows 

 a differential yielding of the strata at this 

 point to the tangential stresses that produced 

 the deformation of the Appalachian province. 

 This brought about transverse faultings with 

 horizontal shearing, buckling, and in the shale 

 around Middlesboro highly confused folding, 

 possibly associated with a local downward 

 flexure of the heavy Lee sandstone. The 

 plain is the result of ponding, and a careful 

 study of all the facts seems to indicate con- 

 clusively that to produce this ponding there 

 has been recent movement along the old fault 

 I)lanes, or folds, or both. This movement has 

 been at least one hundred and possibly several 

 hundred feet vertical, and is possibly post- 

 Tertiary in age. 



The next paper, entitled ' The Significance 

 of U-shaped Glacier and Stream Channels,' 

 was by Prancois E. Mathes. 



The tendency to assume a U-shaped cross- 

 section is not characteristic of glacier chan- 

 nels alone, and should not be looked upon as 

 the peculiar result of ice action. Channels 

 produced by streams of water exhibit the same 

 tendency, and this type of cross-section should, 

 therefore, be considered characteristic of all 

 stream-worn channels, using the term stream 

 in its broadest sense. 



Observation on irrigation canals and ditches 

 teaches that whatever their original cross-sec- 

 tions may be, they will in time be replaced by 

 U-shaped ones. The transformation may be 

 affected: (1) by enlarging, that is by cutting 

 alone, (2) by cutting and filling combined, 

 or (3) by filling alone. The resultant figure 

 is the same in each case, provided the volume 

 and the slope are the same. It further ap- 

 pears that after certain definite proportions of 

 outline have been reached the cross-section no 

 longer tends to change but remains virtually 

 stable so long as the conditions of flow remain 

 unchanged. It is inferred from this that a 

 stream of constant volume, flowing on a uni- 

 form slope, tends to evolve a cross-section of 

 certain definite proportions, this cross-section 

 being the one through which the stream can 

 flow with the greatest economy, that is to say. 



