TJie Theory of Immersion. 



253 



ray itself arrives — we find that, omitting fractious, what we actually 

 get for use is nothing! pe.r cent. The actual proportion is about 

 yIs, which as the original ray itself is so small, is a quantity 

 w^holly imperceptible. 



The other table is for an object mounted in balsam, or any 

 medium of the same kind, the index being here assumed the same 

 as for glass. 



OliJECT IN A MeDHM like BaLSAM. 



Two reflexions only are here encountered. The true apertures 

 are given in the first column, but as these would have to be taken 

 by some special method, as by the " tank " or some equivalent 

 arrangement, the corresponding air-apertures are added for con- 

 venience. The apertures here selected are not in round numbers, 

 because in that case the apparent apertures could not be expressed 

 in round numbers. But if we are proceeding by steps, it is, of course, 

 of no consequence at what exact places we stop. For uniformity, 

 therefore, the round numbers are still kept in the air-angles, thou^-h 

 these, of course, are not the real apertures. And in the table a 

 decimal place is added on account of the smallness of the differences 

 in the higher angles. 



These figures afibrd us all that is necessary theoretically for 

 testing the question whether the superiority claimed for the im- 

 mersion glasses is due to the light lost in passing from medium to 

 medium. I do not profess at present absolutely to answer it, partly 

 for the reasons given above, but partly also because the observations 

 must be made in a special way ; and my principal purpose is to 

 call the attention of those who may possess the requisite skill and 

 leisure and appliances to the principles upon which, whenever it is 

 answered, the answer must be founded. At first sight there seems 

 to be less difference than we might have expected. But then it is 

 to be observed that the numbers, taken in the form in which they 

 stand, do not give us the necessary tests. They must be subjected 

 to a little manipulation first. Take, for example, as an extreme 



u 2 



