1813.] On the Measurement of Minute Particles. 191 



edge of the aperture project so far beyond the other, as to limit 

 its visible extent. Now the substance, in which the lens of a 

 microscope is contained, presents a small aperture capable of 

 exhibiting effects of this kind, which, however, can only be 

 expected to appear when the light is peculiarly circumstanced. 

 The aperture of the highest magnifier that I have employed is 

 ■gL- of an inch, which answers to about n. 330 of the scale of the 

 criometer, and would consequently exhibit a bright ring at -^-^ 

 of the distance of a minute object viewed through it, while the 

 darkest part within this ring would be at about f- of that dis- 

 tance: and the focal distance of the lens being about -^ of an 

 inch, the diameter of the apparent dark circle would be - t , \ . f0 

 of an inch, and that of the bright one f^-fi and the dimen- 

 sions would be nearly the same if any other small lens were 

 employed, with an aperture half as great as its focal distance; so 

 that the constancy of such an appearance, notwithstanding a 

 change of magnifiers, might increase the probability of error. 

 It is obvious that a shade of this kind, surrounding the central 

 parts of a globule, if they happened to be much brighter than 

 the rest, might give rise to a mistaken idea of inequalities in its 

 form or structure ; and it is possible that when a particle is darker 

 than the surrounding medium, some parts of its surface mav 

 have lines of a similar nature projected on them in an inverse 

 order. The particles of the blood are about ifife, of an inch in 

 diameter, varying from d „'„ u to voVoJ an d it is extremely 

 possible that an object of these dimensions may exhibit a light 

 point near its centre, which may be surrounded by a dark and 

 then by a light annular shade within the limits of its disc. There 

 are also several other sources of error in different lights, and in 

 a focus more or less imperfectly adjusted; it is, however, suffi- 

 ciently evident that no fallacy of this kind can have given rise 

 to all the appearances which have been already described, as 

 observable in the particles of the human blood, and still less to 

 those which are observable in the blood of some other animals. 



VI. Changeable Colours. 



In examining some of the dust of the lycoperdon, I had put 

 it with a drop of water on a glass, when I observed a purple 

 ige in the water, which I thought at first was a stain extracted 

 from the- powdei ; but the water viewed separately was perfectly 

 transparent, ami the light transmitted directly through the water, 

 when the globules were present, was of a yellowish green. After 

 some consideration, I conjectured that this appearance of colour 

 must be analogous to that of the mixed plates which I had for- 

 merly observed, depending on the difference of refractive density 



Ihc water and the globules, (Nat. Phil. I. 470, PI. 30, F. 4-30. 

 J J. 633.) and by substituting fluids of different densities for 



