386 THE AMERICAN MONTHLY * [Dec, 



lengths obtained by the use of the ammonio-sulphate of 

 copper solution are recognized as supplying the purest 

 mono-chromatic illumination so called. But the gist of 

 the matter lies in the remaining arrangements. A cir- 

 cular flame, with a convex surface therefore, is used in 

 place of a flat flame, then the bull's eye lens is put as 

 close as possible to the flame, whereas it should be at its 

 focal length, the beam now of other than parallel rays 

 is received on the concave mirror and sent through the 

 achromatic condenser in what kind of way ? Finally to 

 add to the complexity of matters, a narrow angled half- 

 inch objective is inverted and used as an eye-piece. 

 What the resultant image is would be hard to say ; but 

 observations based on such violations of optical prin- 

 ciples governing the obtaining of critical images, can 

 surely not be received as sound. As personal bias can- 

 not, for a moment come into consideration, Mr. Cunning- 

 ham will doubtless be glad to refresh his memory with 

 the following quotation from DalHnger's revision of Car- 

 penter, p 357 : But the supreme folly of using a concave 

 mirror and a bull's eye is shown in figure 310".... 

 this secures a result — as will be seen by the relation of 

 the light to the condenser which is as far from what is 

 sought and desirable as it can well be. While another 

 lesson of great importance may be learnt from fig. 311, 

 which illustrates the error of not having the edge of the 

 flame (e) in the principal focus of the bull's eye (b)." 



"The above are fundamental principles of illumina- 

 tion, and if the student is to succeed as a manipulator 

 he must demonstrate and redemonstrate them." Far- 

 ther, p 362 ; " Mr. Nelson has been able to obtain the 

 most wonderful results with narrow cones, 'true ghosts' 

 and 'false ghosts'. . . .and many complex and false im- 

 ages with the coarser diatoms. But with wide cones he 

 has proved that these false images cannot be produced." 



We conclude, then, that Mr. Cunningham's asserted 



