166 



SCIENCE. 



LxN. S. Vol. XXIY. No. 606. 



The view here advanced, it is thought, 

 presents far less difficulty. It is indeed 

 an extension of Mallet's idea to include 

 other factors of energy besides pressure. 



The fact that the heated volcanic masses 

 are in the earth, surrounded by materials 

 of low heat-conducting power, makes the 

 accumulation or retention of heat energy 

 generated by mechanical work possible. 

 The local character, as well as the varied 

 nature of volcanic phenomena, is not diffi- 

 cult to understand. Volcanic outbursts 

 must continue so long as readjustments of 

 the positions of rock masses under a critical 

 condition of pressure occur with the deeper 

 surface layers. 



If, as appears probable, the earth's in- 

 terior is metallic iron, surmounted by a 

 covering of oxidized lighter material, slag- 

 like in character but altered by water and 

 sedimentation, etc., the interior tempera- 

 ture or that of the metallic body is not 

 likely to be very high, and it must be fairly 

 uniform in spite of the gradual increase 

 noted in the surface rocks at increasing 

 depths. Such surface rocks or layers, being 

 of relatively low heat conductivity, serve, 

 so to speak, as a non-conducting blanket in 

 which alone a considerable temperature 

 gradient would be manifested. This con- 

 dition does not forbid the possibility of the 

 readjustments of masses herein regarded as 

 the initial cause of superheating, even if 

 those masses be quite hot but under such 

 gravitational pressure of superincumbent 

 masses as to forbid fusion. 



Elihu Thomson. 



TEE GEOGRAPHICAL DISTRIBUTION OF THE 

 STUDENT BODY AT A NUMBER OF 

 EASTERN AND WESTERN UNI- 

 VERSITIES AND EASTERN 

 COLLEGES. 



The accompanying table explains the 

 geographical distribution of the student 

 body of six of the leading universities of 



the east and of four western institutions, 

 as well as of three New England and two 

 Pennsylvania colleges for the academic year 

 1905-06, summer session students being in 

 every instance omitted. It was impossible 

 to secure accurate figures for the academic 

 year just closed in the case of the Univer- 

 sity of California, and consequently the 

 figures for 1904^05 were substituted. Com- 

 paring the attendance by divisions of the 

 six eastern universities {Columbia, Cornell, 

 Harvard, Pennsylvania, Princeton, Yale) 

 with the corresponding figures for the same 

 universities included in a similar table 

 published in Science, N. S. Vol. XXII., 

 No. 562 (October 6, 1905), pp. 425-6, we 

 note in the first place that there has been 

 a gain for these universities taken as a 

 whole in every division, the largest increase 

 in actual number of students, leaving the 

 north Atlantic division— in which all of 

 these six universities are located— out of 

 consideration, having been recorded in 

 foreign countries, where there has been a 

 gain of eighty-seven students, this being as 

 large as the entire increase of the student 

 clientele of these universities in the United 

 States divisions (exclusive of the north At- 

 lantic). In the south Atlantic states and 

 in the insular possessions these eastern uni- 

 versities have made only a slight gain; in 

 the south central division all of them 

 show an increase with the exception of 

 Princeton; in the north central division 

 the chief gains have been made by Colum- 

 hia, Princeton and Yale, in the western 

 division by Columbia, Pennsylvania and 

 Yale, and in foreign countries by Column' 

 hia, Cornell and Harvard. These figures 

 bear out the statement made by the writer 

 in earlier contributions to Science to the 

 effect that the western and southern 

 clientele of the prominent eastern uni- 

 versities is not suffering any shrinkage. 

 At Columbia the attendance from outside 



