16 Transactions of the Society. 



It will be interesting now to consider why the second order 

 image happens to be so conspicuous. If it were simply a question 

 of size, one would expect the first order image of the speculum 

 globule to be more conspicuous still, but, in fact, that image can- 

 not be seen. The reason can easily be assigned. The speculum 

 globule lies in a perfectly dark field, and is illuminated only by 

 light which it reflects from the stage of the Microscope. Only its 

 reflecting surface, therefore, is a visible object at all, and thus the 

 image of the speculum globule, theoretically present in the object 

 globule, is an invisibly dark object. Under very special conditions 

 of illumination it can just be seen. But to bring it into view is a 

 difficult experiment, and one which I have not attempted this 

 evening to demonstrate. The images of the speculum globule 

 being thus excluded, we, nevertheless, have to consider a whole 

 series of images of the second, fourth, sixth, etc., orders. These 

 may all be dealt with in a word by considering only the case of 

 the fourth order image. It will at once be appreciated that this, 

 having undergone four reductions in scale by reflections between 

 the globules, has become an object almost infinitesimally small. 

 In fact it is much too small to be seen, and therefore, of all the 

 images which are theoretically possible, only this second order 

 image of the object globule is, in fact, a visible image. It, how- 

 ever, shines with such effulgence as to constitute it a most striking 

 object, very easily identified and observed. Here, then, we have 

 the primary conditions of a perfect test object: A circular disk 

 which is densely black and of known, that is to say, of calculable 

 dimensions, lying in a bright field and capable by a proper selec- 

 tion of mercury globules, of being made to any desired size, so that 

 we can overpass the resolving power of any imaginable lens. 



The optical system built up in this way of the two mercury 

 globules mounted one at each end of the Microscope, has some 

 interesting properties. It is to be observed that the two globules 

 do not occupy positions which are conjugate to one another. On 

 the contrary, each occupies what is an apertural plane in the 

 optical system which focuses in the other globule. Consequently, 

 the two principal focal planes of the globules are conjugate to one 

 another, not their two centres. From this it follows that the 

 dimensions of the image seen depend simply on the principal focal 

 lengths involved, and are independent of the exact positions which 

 the globules occupy. This fact is highly convenient, since it 

 enables us to place the speculum globule at whatever distance 

 from the eye lens is most convenient for the observation that we 

 wish to make. The position of the object globule is, of course, 

 definitely determined, since its principal focal plane must coincide 

 with the focal plane upon the stage of the Microscope. But it 

 may be desirable to vary the position of the speculum globule. A 



