Febeuaey 26, 1897.] 



SCIENCE. 



37 



and dunes to ground and underground 

 water is fully discussed. 



THE PEEGLACIAL KANAWHA. 



The effect of drift obstructions in alter- 

 ing the courses of rivers in western Pennsyl- 

 vania and eastern Ohio has been recognized 

 for a number of years, the northward dis- 

 charge to Lake Erie having been thereby 

 greatly decreased. The diversion of the 

 Missouri from a northward discharge to its 

 present membership in the Mississippi sys- 

 tem is also credited to glacial obstruction, 

 and the important service of the great 

 western river as a guide to western explo- 

 ration, early and late, on our side of the 

 Canadian boundary may, therefore, be 

 credited, along with the water powers of 

 INew England, to the glacial period. Still 

 another example of this kind is noted in the 

 17th Annual Eeport of the Geological Sur- 

 vey, in which the Director mentions a dis- 

 covery by F. Leverett, with reference to 

 the ancient drainage of the Virginias. The 

 Kanawha, uniting with other streams in 

 the western part of West Virginia and east- 

 ern Kentucky, ran in preglacial time 

 northward towards Lake Erie, along a line 

 partly coincident with the course of the 

 south-flowing Scioto of to-daj^. This makes 

 the preglacial drainage of the St. Lawrence 

 include headwaters in North Carolina. The 

 existing Ohio can, therefore, no longer be 

 interpretated as of ancient origin, as if still 

 flowing along a consequent course between 

 paleozoic uplifts on the north and south- 

 east. It is a composite stream of post- 

 glacial date. As a glacial product, it has 

 been of even gi-eater service than the 

 Missouri, for our early settlers in its fertile 

 lower valley took great advantage of its 

 well-graded course, along which their ad- 

 vance was much easier than if they had 

 had to go up and down hill, across the 

 grain of various north-flowing rivers.* 



*Some of my correspondents have pointed out that 



THE RIVERS OF SAGINAW BAY. 



A NUMBER of years ago Gilbert described 

 the course of the Maumee river in northern 

 Ohio, showing that its peculiar back-handed 

 branches were consequent upon the faint 

 relief determined by moraines and glacial 

 lake beaches. A recent essay by Taylor 

 (Correlation of Erie-Huron beaches with 

 outlets and moraines in southeastern 

 Michigan, Bull. Geol. Soc. Amer., VIII., 

 1897, 31-58) now gives another example of 

 a very similar kind and warrants the rec- 

 ognition of these back-handed branches 

 under some appropriate name, ready for 

 convenient use when still other members of 

 the class shall be discovered. Saginaw 

 river, with its Cass and Tittibawasee arms, 

 and swampy head opposite the upper 

 course of Grand river, repeats the essen- 

 tial features of the Maumee to a nicety. 



Back-handed branches resulting from 

 the migration of divides quite independ- 

 ently of glacial constraint are easily distin- 

 guished from the class here considered. 

 The barbed arrangement of the upper 

 branches of the Maira recently diverted 

 from the Inn on the watershed of the Alps 

 are of this second class. 



W. M. Davis. 



Harvard University. 



CUBBENT NOTES ON METEOBOLOGY. 

 CHALK-PLATE WEATHER MAPS. 



One of the recent improvements in the 

 methods used by our Weather Bureau de- 

 serves mention in these Notes. As is gen- 

 erally known, the daily weather maps issued 

 from the various Weather Bureau stations 

 over the country have for some years been 

 reproduced by a stencil process which, al- 

 though a good method, when carefully ex- 



a suggestion regarding the origin of Teay valley 

 (Science, II., 1895, 40) independent of the Kanawha, 

 is inadmissible ; for the valley contains gravels that 

 could only have come from the upper KanaTvha. Its 

 origin by Big Coal river, therefore, seems out of the 

 question. 



