142 TESTING THE MICEOSCOPE. 



We do not, however, obtain in this way (nor in general with 

 similar observations with the naked eye) such harmonious results 

 as in vision through the Microscope. Measurements which we 

 often repeated gave, for instance, with a vision of 200 mm., values 

 for D fluctuating between 100 and 128 mic., increasing re- 

 spectively to 125 and 160 mic., if we suppose the image to be 

 projected at the conventional distance of 250 mm., for which 

 magnifying powers are usually calculated. 



Otto Mueller 1 obtained 103 mic. for his eye as the mean 

 diameter of the meshes, Hart-ing 2 99 mic., and Helmholtz 3 places 

 the limit of discrimination with a trellis-work even as low as 

 78 mic. 4 



Under these circumstances a graphic representation is best 

 adapted to exhibit the results of the calculation, since the con- 

 struction remains the same whatever may be the amplification 

 selected as the starting point. We do not, therefore, give a 

 tabulated view of these quantities, but have drawn the magnifying 

 powers in Fig. 88 according to an entirely optional scale for the 



abscissae, and the corresponding values of - according to another 



* 



1 Cf. " Untersuch. neuerer Mikroskop-Objective," p. 11 (259). 



2 " Mikroskop," 2nd ed. i. p. 81. The diameter of the meshes and the 

 thickness of the wire must, of course, be added, which Harting in another 

 place (i. p. 342) has overlooked. The methods of observation of Harting, 

 moreover, give figures which cannot be compared with those obtained with the 

 Microscope. 



3 " Physiologische Optik," p. 218. The author gives 63'7- f > seconds, which 

 is about equal to the above value. Helmholtz employed for his observations a 

 grating of black wires, their interspaces being equal to the diameter of the 

 wires, and placed them under a bright sky. The figures thus obtained 

 evidently do not admit of comparison with those obtained by our method. 



4 How little such observations in general agree is shown by the table given 

 by Harting ("Mikroskop," i. p. 72), in which the visual angles of the meshes 

 at the limits of discrimination for five observers are tabulated. We consider 

 the methods of observation upon which it is based (according to which the 

 wire-gauze is viewed direct, and the object-distance is increased to three 

 metres), to be inapplicable in the given case, since we cannot assume that the 

 eye always accommodates itself equally well to different distances. At any 

 rate, measurements of dioptric images are to be preferred for comparison with 

 microscopic observations, if they are made with as favourable a distance of the 

 eye as possible. The minute image on the retina depends, therefore, in the one 

 case, as well as in the other, on approximately equal conditions, and the reduc- 

 tion to the conventional distance of 25 cm. is merely a matter of calculation. 



