January 24, 1902.] 



SCIENCE. 



153 



siphon tube, both legs of which are 10 cm. or 

 15 cm. longer than the barometric column. 

 The bore of the tube should be small (about 

 j\ sq. mm.) to work well. Let one of the legs, 

 BC, dip down into a larger tube CD, partly 

 filled to D with mercury. Fill ABC with 

 mercury, and start the siphon drawing mer- 

 cury from C over to A in the usual way. In 

 order to start the siphon the vertical height of 

 B above the surface D of the mercury should 

 be less than the length of a mercury barometer 

 column, but as the flow continues, the mercury 

 surface descends and keeps on descending 

 until its vertical distance below is consider- 

 ably greater than this length. 



/^^ 



S) 



o 



'Jr 



To make this experiment work sufficiently 

 well for demonstration purposes, excessive care 

 in purifying the mercury and cleaning the 

 glass is not necessary. Boiling the mercury 

 in the actual tubes used, for instance, is super- 

 fluous. With ordinary redistilled commercial 

 mercury and tubes cleaned with alcohol the 

 writer has made the siphon work to a height 

 of 70 cm. As the altitude of the University 

 laboratory, where the experiment was per- 

 formed, is a little over one mile, and the ba- 

 rometer pressure, therefore, only aboiit 61 cm., 

 this means that the siphon worked 9 cm. above 

 the barometric height. 



The most plausible explanation of the above 

 fact is that the atmospheric pressure is not 

 the only force pushing the mercury up the 

 shorter leg. It is drawn up partly by the 

 cohesive attraction of parts of the mercury 

 for each other, and the column is kept from 



dwindling by the adiiesive force exerted by the 

 sides of the tube on the mercury. 



It follows from the above that if a mercury 

 siphon is placed under the receiver of an air 

 pump, it can be made to work over a height 

 of several centimeters, even though the air 

 pressure is reduced to only a few millimeters. 

 This experiment also has been shown to the 

 writer's students. The apparatus was similar 

 to that described above, except that the tubes 

 were much shorter. William Duane. 



Hale Physical Laboratory, 

 University of Colorado. 



fossil shells of the john day region. 



Since the ijublication about a year ago* of 

 my paper on the 'Fossil Land Shells of the 

 John Day Region,' etc., I have received from 

 Professor John C. Merriam, of the University 

 of California, a small collection of moUuscan 

 remains obtained by him in the same general 

 locality. Professor Merriam's collection in- 

 cludes examples of the several species of land 

 shells heretofore described,! namely, Epiphrag- 

 mophora fidelis aniicedens, Polygyra Dalli, 

 Ammonitella Yatesi prwcursor and Pyra- 

 tnidula perspectiva simillima. Of these four 

 species there are numerous specimens and 

 fragments. Dr. White's TJnio Condoni appar- 

 ently escaped detection. The foregoing repre- 

 sent all of the molluscan forms thus far re- 

 ported from the John Day beds. Dr. White 

 received his material from the late Professor 

 E. D. Cope and Professor Thomas Condon, of 

 the University of Oregon. Cope's specimens 

 were obtained by Mr. Jacob L. Wortman, of 

 the Army Medical Museum. These two col- 

 lections included the same species. 



Professor Merriam has made some interest- 

 ing additions to the above brief list which are 

 described below. 



helix (ePIPHRAGMOPHORA ?) DUBIOSA NOM. PROV. 



Shell orbicular, flattened, discoidal, peri- 

 phery angulated or obtusely carinated; whorls 



* Proc. Washington Acad. Science, Vol. II., Dec. 

 28, 1900, pp. 651-658, pi. XXXV. 



t Vide Dr. Charles A. White's paper ' On 

 Marine Eocene, Fresh Water Miocene and other 

 Fossil Mollusca of Western North America'; 

 Bulletin No. 18, U. S. Geol. Survey, Washington, 

 1885, with two plates. 



