72 CARNEGIE INSTITUTION 



Students of promise ; and I should greatly regret it if the undevel- 

 oped position of scientific geography resulted in its being set aside 

 for the support of other subjects already so well endowed that large 

 and influential committees of eminent experts are easily organized 

 in their advocacy-. 



The two chief divisions of geography may be further subdivided. 

 Physiography includes all the facts of physical environment of life 

 on the earth ; and the other half of the subject — unnamed, but fairly 

 designated by such a terra as ontography — includes all the kinds of 

 responses of phj^sically environed organisms. Physiography is fur- 

 ther divided into the earth as a globe, the atmosphere, the oceans, 

 and the lands ; ontography may also certainly be subdivided, but no 

 divisions are at present agreed upon among geographers. 



The Earth as a Globe. — All problems concerned with the size and 

 form of the earth are best left with governmental surveys. If par- 

 ticular problems arise in this connection, such, for example, as local 

 magnetic surveys, they will be tluly considered when applications 

 are made by specialists. For the present, no recommendation is 

 made under this heading. 



The Atmosphere. — A. There is mucii need of a scientific study of 

 the failures of weather prediction. Within the last month there 

 have been two examples of failures of precisely the same class as 

 those of tv/ent}^ years ago ; and j-et it does not appear that serious 

 research is in progress with a view to lessening such failures, iduch 

 might be done by a competent investigator who should review maps 

 and predictions and cla.ssify successes and failures. 



B. Another line of work toward the same end is a study of the 

 movement of the upper clouds by means of the horizontal mirror. 

 A study of this kind might be conducted at a moderate expense and 

 useful results could be expected in a year or two. 



C. Detail of weather phenomena. The meshes of the Weather 

 Bureau net are so coarse that many smaller phenomena slip through 

 them undetected. It has frequently been suggested that a scheme 

 of observation by numerous voluntary observers, coordinated under 

 state weather services, might be organized, so as to obtain a closer 

 view of passing changes. The work done in this direction some 

 fi.fteen 3^ears ago by the New England Meteorological Society, and 

 continued for a season or two elsewhere, is well worth systematic 

 extension over a larger area. It is to be expected that the Weather 

 Bureau would permit the use of franked envelopes in such work ; 

 there would be a possibly large expense in providing instruments 



