An introduction to statistics

---

540

JOURNAL OF CHEMICAL EDUCATION

lar fields are inadequate-hut these individuals are not the ones for whom the book was written. In addition to the technical discussions, which in general are excellent, there are chapters on the Relation of Probability to Physical Evidence, Trsjning and Extent of Criminalistics, and The Expert Witness, all of which are interesting and valuable. The advice far the expert witness, if followed, will save him much embarrassment. To the list of subjects recommended as necessary training the reviewer would add a comprehensive course on applied microscopy, since the microscope is the most indispensable tool of the scientific mime investigator. The author does stress the importance of microscopy. His chapters on Equipment for Investigation and The Laboratory, Its Design and Operation will be very helpful to those who are starting new laboratories not only because of the valuable advice but also as substantial argument for adequate financial support for equipment and maintenance. State and muneipitlities should not undertake to develop aueh laboratories unless they me prepared to give continued financial support, not only for equipment but, more important, for the payment of salaries whieh will attract and hold competent investigators. The range of subjects covered is so large that even an enumeration of them is impractical here. The typography is excellent, with but few typographical errors, the illustrations are good, and the book is very readable. I t can be read with profit and understanding by anyone who has s. speaking acquaintance with the basic sciences. The reviewer finds little to criticize since he realizes that the book is for the uninitiated. He does, however, take exception t o one statement where, in discussing the dermal nitrate test, the author ends a fairly good discussion of the earlier farm of the test with the statement that, "It must be remembered that washing the hands before applying the paraffin gauntlet (sic) will effectively remove any test which would otherwise have been found." As a matter of fact repeated washings over a period of at least three weeks do not necessarily remove the particles of nitrate embedded in the skin. Furthermore, the teat is made more reliable if tho hands are washed before the paraffin is applied, fiinoe this will remove surface nitrates that may be present from other caums and whose presence might interfere with the true nitrate tent which is concerned with particles that have become emhedded in the skin because of the force of the explosion. However, among the thousands of true statements made in this excellent book a t least one questionable one should bepermitted. J. HOWARD MATHEWS

0

ORGANIC CHEMISTRY

Ray Q. Brewster, Professor of Chemistry, University of Kansas. Second edition. Prentice-Hall, Inc., New York, 1953. xii 855 pp. 29 figs. 45 tables. 16 X 23.5 cm. $7.

+

THrs is B text for a full-year course. It treats and aromatic compounds separately. The first edition of Brewstcr's text gained wide recognition for its extensive reference to mechanisms in discussing fundamental It was a pioneer, entering an area in which therewere fen examples to follow, in the pitfalls were not mrtrked, The number of its followers is a tribute to the quality of the trail which it blazed. stagehas heen passed, Nous, five years later, the N~~ it is time, though we may he pleased by the general direction of the trail, to expect that some of its rough spots be smoothed, its hairpin turnsbe and its steep gades be moderated. Unfortunately Brewster offers, in this new edition, essentisllv the same oioneer t,rail. More specificslly, one would have wished that the somewhat awkmrd expositions of certain theoretical points had been replaced by smoother, less complicated discussions; indeed, some

passages even tend to give erroneous impressions (for example, the discussion of nucleophilic displacements in alkyl halides, page 96, infers that all such reactions proceed by the 8 9 mechanism). Again, one would have wished that a better balance of emphasis between mechanisms and classical organic chemistry had been attained; for example, the importance of halogen addition to alefins in preparative and analytical work continues to he obscured by a long-winded disoussion of the mechanism of such reactions. Further, one would have wished that modern concepts of reaction mechanisms had beenintroduced in certain chapters which are rather old-fashioned; for example, chapter 25 (Aromatic Nitro Compounds) fails to mention the well-established mechanisms of aromatic nitration and fails, in connection with reactions of chloronitrobeneenes with ammonia, sodium ethoxide, etc., ta identify these reactants as nucleophilic reagents, though this term is used extensively earlier in the hook. A modern point of view is also lacking in parts of chapter 20 (Amino Acids and Proteins). The valuable and significant technique of enzymatic resolution of amino acids is not mentioned, modern methods for the synthesis of peptides do not appear, and the chapter fails to convey the enormous importance of proteins in life processes and the enormous difficulties of working with them. Certain features of the new edition represent a distinct advance aver the old. There are approximately twice ss many problems a t the end of each chapter. A chapter an organometallic eompounds has been added. And in numerous places small changes in content or phrasing have been made. No textbook is perfect. The teacher always h d s i t necessary t o add here, t o omit there. Those \\.ha found the first edition of Brewater's text convenient to teach from will like the new edition better. Those who tried the first edition and then looked elsewhere will probably not be attracted back bythe second edition. Uxrveasmr ol.N o m n C ~ n o b r m C"*FEG HILL,NORT"C*ROLIK*

0

.

JOSEPH F. B U N N E T T

AN INTRODUCTION TO STATISTICS

Chclrles E. Clark, Associate Professor of Mathematics, Emory University. John Wiley 8 Sons, Inc., New York, 1953. x 266 pp. Illustrated. I5 X 23.5 cm. $4.25.

+

THISbook is an elementary introduction to modern methods of statistical inference from samples, intended for workers in any field in whieh these methods may be applied. The author intends to presuppose a minimum of mathematical training and maturity, and in fact scrupulously minimizes even elementary algebraic manipulations, and touches no concept of calculus, or of snslyties, except that understanding of Cartesian coordinates is assumed. The expository style is, however, conventionally mathematical and seems to this reviewer to lack a corresponding forbearance. Abundant use of letter svrnbols, references to pfevious equations, theorems, sections, and the like, may well gwe a mathematically inexperienced reader an unu-srrmted feeling of being in too deep water. The subject matter of the book, however, is very well chosen and amply illustrated by numerical examples. There is a brief introductory discussion of permutations and combinations, and probability concepts. The main portion of the text treats ttests, analysis of variance for s. single attribute, and chi-square testa, with clear emphasis on the underlying concepts of sampling distributions and confidence limits. The final chapter deals in a simple way with product-moment correlation and linear regression. Much problem material is included; the ansuwrs and solutions for odd-numbered exercises alone cover 34 pages. Six numerical tables make the hook self-sufficient in this regard.

,,s

PHILIP RUDNICK INBTITOTION

OCE~N~(IR*PXY

LA JOGLA,c ~ ~ m o n a r ~