388 BELL SYSTEM TECHNICAL JOURNAL 



usual conception of high-power metallography seems to be great 

 enlargement, indistinct detinition and lack of resolution. Such 

 results, generally, have been classed under the heading "empty 

 magnification" because they have failed to show more detail than has 

 been shown at lower magnifications and with objectix-es of less re- 

 sohing ability. Oftentimes the pictures would be luiintelligible 

 taken by themselves, but the reason they are recognized at all is 

 because the same structures have been seen and identified by low- 

 er medium-power methods. Such high-power results are like an 

 elastic band which has been stretched unduly. As the band is stretched 

 it becomes more and more attenuated and finally snaps. If the 

 optical image is stretched In' enlargement the details of the image 

 become less and less distinct and finally the image breaks down alto- 

 gether, so that the detail and the background blend together into a 

 hazy outline of what formerly was a sharp image. 



High-power metallography as presented in this article consists of so 

 preparing metallurgical specimens that crisp, brilliant images may be 

 obtained and photographed at high powers and of achieving approxi- 

 mately the potential resolving possibilities of splendid objectives. 



By improvements in the method of preparing metallurgical speci- 

 mens and in the technique of manipulating the apparatus, "empty 

 magnification" is no longer synonymous with high-power photo- 

 micrography. 



It is the object of this contribution to show the application of this 

 new tool for metallurgical research to the study of metal structures 

 which heretofore have not been resolved and the nature of which 

 has led to much speculation and to wide dififerences of opinion. A 

 clear understanding of the current conceptions of magnification 

 and resolution is essential and a knowledge of the limitations which 

 were regarded for many years as restricting the employment of high 

 powers will prove of value in the interpretation of the results obtained. 

 For this reason a brief discussion follows which not only shows the 

 method of approach in the present development, but indicates the 

 path along which we may work to secure a higher order of resolution. 

 By resolution is meant that property of a lens system which enables 

 it to distinguish or "resolve" as separate and distinct units fine struc- 

 tural details spaced very close together. 



Research and Engineering} E. B. Cr.\ft. Research in industry — 

 which the author mentions is of comparati\ely recent origin — is 

 defined as the application of methods of systematic and logical dcduc- 



I Address before the Engineers' Club, Pliila., Oct., 1925. Engs. and Engg., Jan 

 1926, Vol. 43, pp. 11-19. 



