E. M. NELSON ON METHODS OF ILLUMINATION. 291 



will be noticed between those obtained with a centred and de- 

 centred mirror-image in the Ramsden disc. The student should 

 remember that it is far better to have a centred illumination, 

 even at the expense of an incompletely lighted field, than an 

 evenly illuminated field and decentred illumination. 



To illustrate this 'subject further, let the flat of the flame of 

 a paraflin lamp be used as an illuminant, then, with concave- 

 mirror illumination improperly arranged by a fumbler, the image 

 in the Ramsden disc would very probably appear as in Fig. 1. 

 The image of the flame resembles a decentred slit of light, not- 

 withstanding that the flat of the flame is presented to the mirror. 



It is this decentring of the illumination which causes the coma 

 to rock upon focal adjustment. It is the asymmetrical arrange- 



sooooo 



ment of the beam passing through the objective which destroys 

 the sharpness of the image, and it is the small W.A. {I.e. too 

 large an unutilised area in the objective) which coats the image 

 with black-and-white diffraction images. This image in the 

 Ramsden disc (Fig. 1) should be compared with Fig. 2, which 

 shows that the concave mirror is in correct adjustment, therefore 

 the flat of the flame is imaged properly and centred to the disc. 

 It does not take a single second of time longer to set the mirror 

 properly than improperly. If a student knows what to do, he 

 can do it, but at the present time he knows nothing whatever 

 about it ; and how should he when his instructor merely tells him 

 to fumble about with the mirror until he gets a fully illuminated 

 field ] 



Fig. 3 illustrates a properly-set-up illumination with the plane 

 mirror. The only difference is that the image of the flat of the 

 flame is smaller, and the unutilised portion of the objective 

 larger ; which is, as we have just seen, disadvantageous. With 



JouRN. Q. M. C, Series II. No. 69. 20 



