850 



SCIENCE 



[N. S. Vol. XXX. No. 780 



which this movement "will have vehen once it is 

 in complete operation. It is a comparatively 

 easy matter now for the colleges to take care 

 of this short-course work and a considerable 

 amount of secondary work, because the num- 

 ber of students so far have been comparatively 

 limited in each state. But as we approach 

 the time when we are to have half a million 

 students in agriculture in secondary schools 

 it is going to be a very different proposition. 

 In the near future the colleges will have 

 all they can do to take care of the students 

 in regular college courses in agi-iculture. 

 The special agricultural schools will fill a 

 great need by attracting the more mature stu- 

 dents who would not go to the ordinary high 

 schools, and the ordinary high schools will 

 have plenty of agricultural students of proper 

 high-school age. 



As I said, I believe the standard courses in 

 these special agricultural schools should not 

 be narrowly vocational, but ghould conform, 

 in a general way, to the general standard for 

 the high-school system in the state, and they 

 should be organized so as to connect them 

 definitely with the general educational system 

 of the state. To do this it will probably be 

 found necessary in the case of schools that 

 have shortened the school year to twenty-four 

 weeks of six days each, instead of thirty-six 

 weeks of five days each, to add another year 

 to the standard course, making it five years 

 instead of four. But it would be desirable 

 that besides the standard courses which would 

 prepare the student for college or for life, as 

 the case might be, such schools should have 

 shorter courses more purely vocational. 



SCIENTIFIC BOOKS 

 Experimentelle Untersuchungen uber Atom- 

 gewichte. Von Theodore William Rich- 

 ards und seinen Mitarbeitern. Berlin, 

 1909. 



In this fine octavo of 890 pages. Professor 

 Richards has brought together, in German 

 translation, the many papers upon atomic 

 weights which, during the past twenty-two 

 years, have been published by him and his 

 collaborators. These researches are already 



well known to all chemists who are interested 

 in the accurate determination of these funda- 

 mental constants, and the results obtained 

 have received very general acceptance. Their 

 collective publication, however, is highly sug- 

 gestive, and deserves a careful review. 



The first of these researches, that upon the 

 atomic weight of oxygen, was carried on by 

 the late Professor J. P. Cooke, with the co- 

 operation of his then student, Richards. The 

 latter began his independent work with a re- 

 vision of the atomic weight of copper, which 

 was followed by papers upon barium and 

 strontium. Afterwards, Professor Richards 

 had the assistance of his advanced students, 

 and with their aid the atomic weights of zinc, 

 magnesium, nickel, cobalt, iron, uranium, 

 calcium, caesium, sodium, chlorine, potassium, 

 nitrogen, sulphur and silver have been rede- 

 termined, and apparently with the greatest 

 possible accuracy. There is no reasonable 

 doubt that the work done has been a great 

 advance upon all previous investigations of 

 similar purport; but as Professor Richards 

 would himself admit, it is neither final nor 

 absolute. Our knowledge of physical con- 

 stants is obtained by what may be called a 

 method of successive approximations; but 

 absolute accuracy is unattainable. The re- 

 searches now before us represent, in all proba- 

 bility, the closest approximations to the truth 

 as yet reached, but that statement does not 

 imply the impossibility of future improve- 

 ment. Such improvements are likely to be 

 small, however, and to affect only the minor 

 decimals. 



In reviewing the work so far done, one can 

 not help noting the steady advance in experi- 

 mental technique. The later determinations 

 appear to be of a much higher order than the 

 earlier ones. Indeed, several of the papers in 

 the volume are devoted to improvements in 

 manipulation, or to the exposure of constant 

 errors against which the investigator must be 

 always on his guard. The bottling apparatus 

 in which materials are prepared for weighing, 

 and the nephelometer by which mere traces of 

 precipitates are recognized, represent im- 

 provements in apparatus. The purification of 



