April 7, lOOo.] 



SCIENCE. 



549 



P. Gratacap, C. S. Hastings, J. P. Iddings, 

 C. L. Jackson, S. P. Johnson, S. C. Keith, 

 N. T. Lawrence, G. Lefevre, C. K. Leith, E. 

 W. McBride, L. B. Mendel, T. C. Mendenhall, 

 A. A. Michelson. John Muir, E. W. Nelson, 

 E. L. Nichols, A. E. Ortmann, Wm. Palmer, 

 H. S. Pritchett, T. M. Prudden. H. A. Purdie, 

 E. F. Smith, J. C. Smock, E. Thaxter, O. H. 

 Tittmann, John Trowbridge, W. L. Under- 

 wood, Lester F. Ward, A. G. Webster. E. L. 

 Wells, C. A. White, S. W. Williston, H. C. 

 Wood and R. R. Wright. It was not practic- 

 able to apply a similar test to the foreign lists, 

 and it may be that they are fuller. 



To test the accuracy of the addresses given 

 they were compared with lists, of approxi- 

 mately the same date, published by the Wash- 

 ington Academy of Sciences, the Geological 

 Society of America, the American Society of 

 Naturalists and the American Ornithologists 

 Union, and with 400 other addresses taken at 

 random from ' Who's Who in America ' and 

 the lists of the American Association and the 

 affiliated societies of Washington. In all about 

 750 addresses were compared, and it was found 

 that about ten per cent, of those given by the 

 directory are erroneous. Similar comparison 

 was made of 291 names common to the direc- 

 tory and the list of the Geological Society of 

 London, with the result that 52 addresses were 

 found to be discrepant, but in this case it was 

 not possible to say how many were wrong. 



Of deceased scientists so notable that their 

 deaths are recorded in the necrologies of the 

 National Academy, the American Journal of 

 Science, or the American ' Who's Who,' no 

 less than 49 are retained by the directory. 

 Among these are not only Powell and Le 

 Conte, already noted, but Elliott Coues. Ho- 

 ratio Hale, James Hall, J. Willard Gibbs, St. 

 George Mivart, Henry Morton, A. E. Nord- 

 enskiold, H. A. Rowland and Rudolph Vir- 

 chow. 



The arrangement of the names is by coun- 

 tries, with a classification which has been 

 gradually evolved through successive editions. 

 Part I. comprises, first, the United States and 

 Canada, and then, in order. Great Britain, 

 Central America, South America, Oceanica 

 and Africa. Central America is made to in- 



clude not only the usual states, but Mexico, 

 Newfoundland and the islands of the West 

 Indies ; and the countries of Asia are placed 

 under Oceanica. Part II. includes all the 

 countries of Europe except Great Britain. 

 The use of Part II. is facilitated by having 

 its parts arranged in alphabetic order, and by 

 the insertion of the name of the country at the 

 head of each page; but these devices are not 

 used in Part I. In some of the earlier edi- 

 tions the entries for the United States and 

 Canada were numbered seriatim and a special 

 index of departments of science referred to 

 these numbers. From the present edition the 

 index is omitted, but the numbers survive as a 

 vestigial character. 



The personal list for the United States and 

 Canada is followed by a list of scientific so- 

 cieties of the same countries, with a classifi- 

 cation by states. Being a resident of Wash- 

 ington, I turned, naturally, to the list for the 

 District of Columbia, and noted at once the 

 omission of the Washington Academy of Sci- 

 ences and of nine out of the twelve scientific 

 societies affiliated with it. Of the three affili- 

 ated societies that are listed the data for two 

 are obsolete. The American Association for 

 the Advancement of Science, which for seven 

 years has had its headquarters in Washington, 

 is still credited to Salem, the place of publica- 

 tion of the directory. 



Despite these limitations the directory is a 

 useful volume. It contains the names of about 

 18,500 scientists, with information as to ad- 

 dresses and specialties, and the greater pan-t of 

 the information is correct. 



G. K. Gilbert. 



AN OVERLOOKED FORM OF STEREOSCOPE. 



In Science of November 18, 1904, Professor 

 Jastrow describes, under the above heading, 

 an ingenious modification of the mirror stereo- 

 scope, permitting the use of the ordinary 

 stereoscopic card. 



The arrangement described below, serving 

 the same purpose, appears to possess some ad- 

 vantages. It is quite possible that this form 

 may have been suggested before, but it has not 

 come under my notice. 



In all arrangements of this kind it is of 



