Maech 13, 1896.] 



SCIENCE. 



411 



and the descriptious are almost entirely inde- 

 pendent of each other without definite com- 

 parisons. 



The third part of the work (184 pp.) is given 

 to his subject proper; it is somewhat unequal 

 in character, being much more detailed and 

 careful in the earlier portion than in the later. 

 Here, too, one looks in vain for comparisons or 

 for any definite reasons for the inclusion of 

 some of the insects in the groups in which they 

 are placed, by references to the earlier portion 

 of his work. The classification is entirely novel 

 and bears little relation to that employed by 

 the present writer, which is an extension of 

 that of Dohrn and Goldenberg. This is not 

 the place for a discussion of the relative merits 

 of the two, which may be left to the impartial 

 student of the future ; but in giving up the term 

 Palseodictyoptera for the bulk of paleozoic in- 

 sects, as indicating the far greater afiiliation of 

 insects in paleozoic time than subsequently, 

 Brongniart overlooks the fact that while his dis- 

 coveries show a wider diversity of forms among 

 paleozoic insects and more definite points of 

 relationship between them and and later types 

 than we have ever had before, they but empha- 

 size and further illustrate the reasons for which 

 the name was proposed. General statements 

 previously made regarding paleozoic insects as 

 a whole are in no way weakened by this great 

 extension of the field, and this renders the im- 

 portance of these generalizations even greater 

 and their validity surer than before. 



The work is most luxuriously issued and the 

 plates all that could be desired, excepting that 

 many of those illustrated by heliogravure (in 

 the most artistic manner, indeed) need to be 

 supplemented by drawings showing the precise 

 origin of each of the veins ; these are often ob- 

 scure in the best photographic picture, since 

 they very often cannot all be seen in any single 

 view, or their contrast to the stone is insufii- 

 cient for clear results. Why the title page 

 should bear the date 1893 is difiicult to under- 

 stand, for the second signature (p. 12) contains 

 a long extract first published in America in 

 February, 1894, and the earliest copies of the 

 work only reached this country in June, 1895. 

 Except in the separate ' Explanation of the 

 Plates ' in the atlas, no reference to the figures 



occurs in the text, which is a great inconven- 

 ience. Samuel H. Scuddee. 



Revision of the Shrews of the American Genera 



^larina and Notiosorex. By C. Hart Mbr- 



EIAM, N. Am. Fauna, No. 10, December 31, 



1895, pp. 5-34, pU. 1-3. 

 The Long-tailed Shrews of the Eastern United 



States. By Gebrit S. Miller, Jr. Ibid., 



pp. 35-56. 

 Synopsis of the American Shrews of the Genus Sorex. 



By C. Hart Meeeiam. Ibid., pp. 57-98, 



pll. 4-12. 



The shrews are among the most difficult of 

 mammals to discriminate specifically, owing to 

 their general similarity in color, size and general 

 external appearance. Hence resort must be 

 had to the teeth, which, though minute, often 

 afford trenchant characters. No group of 

 American mammals has hitherto been in a more 

 thoroughlyunsatisfactory state, as regards either 

 the number and distribution of the species or 

 the names they should properly bear. Hence 

 the three papers on the American Shrews that 

 constitute No. 10 of ,' North American Fauna ' 

 are a particularly welcome contribution to the 

 literature of North American mammalogy. Two 

 of these papers are by Dr. C. Hart Merriam and 

 the other is by Gerrit S. Miller, Jr., and jointly 

 they comprise a careful revision of the whole 

 group. The work is based primarily on the col- 

 lections brought together by Dr. Merriam under 

 the auspices of the United States Department of 

 Agriculture, the only outside material used be- 

 ing mainly the type specimens of previous 

 authors, which in most cases have been acces- 

 sible to the authors of the papers under notice. 



Formerly shrews were rare in collections ; 

 generally they were so difiicult to obtain that 

 only chance specimens were secured. That 

 such is no longer the case is evident from the 

 large number of specimens now accessible for 

 study in most large collections of mammals, 

 very successful methods of trapping these ob- 

 scure and mainly nocturnal animals having 

 been discovered within comparatively recent 

 years. Thus the Department of Agriculture 

 collection alone numbers upwards of 2,000 

 specimens, brought together largely within the 

 last six or eight years. 



