Januaet 29, 1897.] 



SCIENCE. 



191 



cross-section must change to produce a smaller 

 flow as we descend the glacier. The ice near 

 the lower end of the glacier is under greater 

 pressure than the normal pressure and would 

 therefore have a tendency to rise. Stratifica- 

 tion is not easy to recognize, but certain obser- 

 vations support the view that this potential ris- 

 ing becomes actual, which accounts for many of 

 the phenomena observed. The surface at any 

 point depends on velocity and rate of melting, 

 and varies inversely with them. The larger 

 the glacier the greater will be the differential 

 motion. For small glaciers the differential 

 motion is small and the slope steep. If any- 

 thing cause an abnormal melting of the lower 

 layers the upper ones will advance but over 

 them, which is thought to be the explanation of 

 certain facts observed in Greenland by Cham- 

 berlin and Salisbury. Although the sloping 

 surface of Alpine glaciers is a surface of equi- 

 librium it is unstable. If the surface be in 

 equilibrium it will respond quickly to climatic 

 variation. If it be widely removed from equi- 

 librium it will respond more slowly, an explana- 

 tion of the variation of glaciers differing from 

 both that of Forel and Richter. 



Loess in the Wisconsin Drift Formation : By R. 

 D. Salisbtjey. Loess has long been known to 

 cover the glacier drift of the earlier epochs at 

 various points. At least two loess sheets are 

 known, one of which is correlated with lowan 

 ice and passes beneath the Wisconsin. Hereto- 

 fore loess has not been known to occur in or 

 above the "Wisconsin drift, but during the past 

 summer it has been found in connection with 

 this formation at several points in Wisconsin, 

 namely, Green Lake, Devil's Lake and Abie- 

 mans. 



Geology of Chiapas, Tabasco and the Peninsula 

 of Yucatan: By Carlos Sapper, translated 

 by C. JoAQUiNA Maury and G. D. Harris. 

 Southeastern Mexico shows three series of 

 formations : an ancient complex of Palaeozoic 

 beds and eruptives in southern Chiapas ; a more 

 modern belt of Mesozoic and Tertiary forma- 

 tions in the middle and northern regions ; and 

 a third zone of great plains at the foot of the 

 other belts, only slightly elevated above sea 

 level, and of Quarternary age. In the penin- 

 sula of Yucatan there is less diversity and the 



beds are undisturbed, in which regard they are 

 contrasted with those of Chiapas. Descriptions 

 of the various formations, with lists of fossils, 

 are appended. 



Studies for Sttidents — Stratified Drift : By R. 

 D. Salisbury. Water must have been an im- 

 portant factor in the deposition of the drift, par- 

 ticularly along the margin of the ice. A much 

 larger amount of drift is stratified than is com- 

 monly thought. These deposits include extragla- 

 cial, supermorainic, submorainic and intermo- 

 rainic stratified drift. The deposits made during 

 the advance, maximum extension and retreat 

 of an ice sheet show certain differences. During 

 maximum extension there was a chance for the 

 development of the following forms : (1) kames 

 and kame belts at the edge of the ice ; (2) flu- 

 vial plains and valley trains in virtual contact 

 with the ice at their heads ; (3) border plains 

 or overwash plains in virtual contact with the 

 ice at their upper edges ; (4) ill-defined patches 

 of stratified drift, coarse or fine, near the ice ; 

 (5) subaqueous overwash plains or deltas 

 formed either in the sea or lakes at or near the 

 edge of the ice ; (6) lacustrine and marine de- 

 posits of other sorts, the material being fur- 

 nished by waters arising from the ice. The 

 same deposits might be formed during the ad- 

 vance of the ice, but would be subject to de- 

 struction by the overriding of the latter. They 

 might be formed during the retreat, but in the 

 latter case the formations dependent upon ice 

 edge would not be so sharply formed. Super- 

 glacial streams are believed to be of only slight 

 importance in this connection, because of their 

 high velocity and the small amount of material 

 upon the surface of the ice. Subglacial streams 

 are considered to be the most probable means of 

 the formation of eskers. H. F. B. 



TERRESTRIAL MAGNETISM, OCTOBER. 



In the first article, by Lieutenant-General de 

 Tillo, entitled Isanomales et Variations Secu- 

 laires des Composantes Y et X de la Force Mag- 

 netique Horizontale pour VEpoque 1857, the au- 

 thor concludes his series of charts of ' isanomo- 

 lous lines ' and of ' lines of equal annual 

 secular variation.' As the title implies, the 

 accompanying four colored plates apply to the 

 westerly component (F) and the northerly com- 



