ULTRAMICROSCOPY 57 



if an object is too small, we can no longer hope to make out its 

 details, however much the magnifying power of the instrument 

 be increased. The resolving power can be increased by using 

 light of shorter wave-length, that is, by using light towards 

 the blue end of the spectrum ; but again a limit is soon reached. 

 Moreover the experimental difficulties are increased. 



In 1903 Siedentopf and Szigmondy invented the ultra- 

 microscope, by the aid of which the existence of particles much 

 smaller than those hitherto known has been demonstrated. 

 The hypotheses of Faraday and others on the nature of 

 coloured glasses, crystals and colloidal liquids are no longer 

 hypotheses, and our knowledge of these and other objects has 

 been greatly extended. The method has also been applied, 

 with advantage, to biological work ; it is probable that, in the 

 near future, its application will be still further extended. 



It will be the aim in this paper to give some account of the 

 ultramicroscope, its development and applications and of the 

 ultramicroscopic particles the existence of which has been 

 revealed by its use. In the few years that have elapsed, a vast 

 amount of work has been done on the subject and it is possible 

 to consider only a few of the many directions in which the 

 use of these methods has thrown some light on the problems 

 of science. 



The Resolving Power of the Microscope 



At the outset, it may be useful to consider the more salient 

 facts with reference to microscopic vision and the resolving 

 power of microscopes. It is well known that a point source 

 of light gives rise to an image consisting not of a single point 

 but something more complex. If the aperture of the objective 

 is circular, the image consists of a circle of light surrounded 

 by diffraction rings, the diameters of the successive rings 

 depending on that of the aperture ; if it is rectangular, we have 

 for the image a rectangular patch of light surrounded by a 

 network of light and darkness. Consider a neighbouring point 

 source. This will give rise to an image similar to that of the 

 former and, if the sources are close together, the diffraction 

 pattern of one may partially overlap that of the other ; when 

 this overlapping exceeds a certain amount, the two images can 

 no longer be separately distinguished. 



It can easily be shown that the smallest distance apart of 



