330 H. M. JOHNSON 



lines always lie in parallel equidistant straight lines (III). If the 

 separation of the lines in systems I and II be equal, the lines III 

 bisect the angle I, II (</>). If one substitute for the lines I 

 and II dark bands of perceptible width (Fig. 2), the lines III 

 become bands of maximal brightness; between two such bands 

 lies one of maximal darkness. These striae are easily per- 

 ceived If the angle <£ between I and II be changed, 



the separation (I) of the striae III is changed accordingly. If 

 4> be small the change in separation is very great, as appears 

 in the relation 



4> c 



Sin — = — 

 2 21 



wherein c represents the separation of the lines 



in I and II. This fact can be utilized in dilatometric demon- 

 strations ; generally in rendering small displacements visible ; and 

 especially in measuring thickness." Behn used the instrument 

 for measuring small changes in expansion, deriving the linear 

 change from change in size of the visible striae; the bar under 

 observation being attached so as to move one of the gratings, 

 the other end of the bar being fixed. I include a reproduction 

 of Behn's figures. 



Ives, 9 who does not refer to Behn's article, restates the relation 

 between the width of the visible striae and the angle of rotation 

 which the former had expressed. In addition to the properties 

 which Behn enumerates, Ives mentions that the mean brightness 

 remains constant through changes in size of the test-bands. 

 (It may be seen from casual inspection of the photomicrographs 

 shown by Ives and by Cobb in the article cited below, that the 

 clear rhombi whose linear arrangement forms the bright striae 

 compose one-fourth the total area of any symmetrical portion 

 of the field.) Ives proposed that the principle be applied to 

 the construction of a test-object for investigation of visual acuity, 

 since no satisfactory instrument then existed and there was a 

 demand for one. He also set forth certain practical considerations 

 governing the choice of grating- width to be used. 10 



9 Ives, H. E., "A Visual Acuity Test-object," Electrical World, lv, 1910, pp. 939 ff. 



10 Ives asserts: " The ideal test-object is one in which the size of the detail is the 

 only variable." ' The test-object as constructed fulfils the requirements of an ideal 

 test-object. The distance of the object and the flux of light entering the eye remain 

 constant. Dark bands, whose visibility forms the test of acuity, may be varied in 

 their separation from invisibility to easy visibility." The present writer believes 

 this valuation of the instrument should be qualified. The instrument fails in one 



