June 11, 1897.] 



SCIENCE. 



923 



to a definite physical phenomenon, has some 

 practical advantages, and it will doubtless re- 

 ceive extended notice and discussion by en- 

 gineers. An authoritative definition of elastic 

 limit will probably be established in time by 

 the international association recently established 

 for the study and unification of methods of 

 testing. 



The author laj's much stress upon the method 

 of judging the quality of a material by means 

 of the work required to rupture it, or by its re- 

 silience, as Thomas Young called it in 1803. 

 The diagram of a tensile test enables this work 

 to be computed, and undoubtedly too great at- 

 tention has heretofore been paid to the ultimate 

 elongation and too little to the ultimate resili- 

 ence. The elongation depends upon the form 

 and length of the specimen and is far from be- 

 ing an absolute measure of the ductility; more- 

 over that part of it which occurs after the maxi- 

 mum strength is reached is of doubtful value in 

 estimating the work of rupture. It is for these 

 reasons that percentage of reduction of area is 

 extensively used in commercial tests, it being 

 found to be nearly independent of the length of 

 the specimen and hence a better index of duc- 

 tility. In this direction of investigation great 

 advances are to be expected, and the develop- 

 ment of impact tests now in progress really re- 

 sults from the desire for a better determination 

 of the ultimate resilience than the static stress 

 diagrams can give. If all tests of metals except 

 one were to be abandoned, the simple test of 

 bending a cold bar by blows of a hammer would, 

 by an overwhelming majority of votes, be the 

 one to be retained ; further, if this cold-bend 

 test be made by a single blow, and if the changes 

 of length on the tensile and compressive order 

 be measured, a determination of both resilience 

 and ductility is obtained, which, though not an 

 absolute one, is probably as valuable as that 

 given by the common static tension test. For 

 these reasons it is thought that the author has 

 somewhat overestimated the value of the ulti- 

 mate elongation as determined on testing ma- 

 chines, and that reliance upon it as an absolute 

 measure of ductility is generally too high. 



The space devoted to the different materials 

 is about as follows : 124 pages on timber, 43 on 

 brick and stone, 77 on cement and mortar, 43 



on cast iron, 24 on wrought iron, 87 on steel, 

 and 18 on alloys. A timely chapter on the 

 magnetic testing of iron and steel, by W. A. 

 Layman, concludes the book. There are over 

 600 illustrations, of which about one-half are 

 the valuable graphic representations and com- 

 parisons. From the extended experience of the 

 author in laboratory work, and from his record 

 as a writer and investigator, it was to have been 

 expected that this book would be an excellent 

 one. It has, however, more than realized the 

 expectations in its Parts III and IV, for here 

 are presented such careful and comprehensive 

 analyses of modern methods and results that 

 the book must at once take high rank as one of 

 the standard authorities on the materials of en- 

 gineering. 



Mansfield Mbrriman. 

 Lehigh University, June 1, 1897. 



Experimental Morphology. Part I. By. Dr. C. 



B. Davenport. The Macmillan Company. 



1897. 



The broadening of the biological horizon in 

 recent years has necessitated an ever-increasing 

 specialization on the part of investigators in 

 that department of science. The territory now 

 open to study is so extensive that it is beyond 

 the powers of any individual to examine all 

 parts of it in detail, and, consequently, each 

 must choose for himself a portion of greater or 

 less extent with which he may expect to be- 

 come tolerably familiar. And yet it is impos- 

 sible to reap the full benefits of results so ob- 

 tained unless they can be correlated with what 

 is being accomplished in adjacent fields, and, 

 that his work may approach the ideal condition 

 of being totus teres atque rotundus, the investi- 

 gator of to-day must look to his neighbors to 

 supply him from time to time with statements 

 of what they have accomplished. Dr. Daven- 

 port's work on Experimental Morphology aims 

 to be a statement of this kind, its object being 

 to review what has been accomplished in the 

 study of the extrinsic forces which determine 

 the course of the development of organisms. 

 The work, as projected, is to consist of four 

 parts, of which the first, now before us, treats 

 of the action of external forces, chemical and 

 physical, on living protoplasm in general, while 



