321 



light reaching this latter till after the two reflections, by which it is 

 confined to one side of the base of the paraboloid and of course 

 reaches the focus and the object at one side only, and by rotating 

 the paraboloid (I believe it ought to be an ellipsoid) and prism to- 

 together, this oblique illumination may be carried round the entire 

 field of the microscope. 



" This arrangement has realized my expectations. Having got 

 it ready in October last, the first object on which I tried its capa- 

 bilities was the Podura scale, an object which has been observed 

 vastly more than any other test. I immediately saw appearances in- 

 dicative of a structure entirely unknown before. In one azimuth of 

 the light the scale appeared obscure and structureless, except that 

 it was studded over with minute, nearly transparent dots, more re- 

 sembling oil-glands in the leaves of the myrtaceee than anything 

 else with which I am acquainted. 



" By rotating the light, faint shadows began to be visible in con- 

 nexion with these dots, and when one-fourth of a revolution or 

 thereabouts was completed, the dots had disappeared and the scale 

 seemed to be covered, thatched as it were, with short, slender, cy- 

 lindrical appendages ; continuing the rotation in the same direction 

 the scale gradually became obscure as before ; these appendages 

 ceased to be visible, and at the end of a second quarter revolution 

 the luminous dots again appeared, more faint than at the opposite azi- 

 muth, still unquestionably there ; completing the revolution, the 

 same succession of appearances recurred until, on returning to the 

 original azimuth, the dots re-appeared as luminous as at first. 



" From frequent repetition of this observation, I am satisfied 

 that this scale, instead of being covered with sculptured lines or 

 folds, as was once believed, or with dark hairs, as figured by Queckett, 

 is in reality covered with cylindrical appendages like quills, but 

 which are either hollow or quite translucent through their axes. 



" I infer this from finding that the several appearances are inva- 

 riably and only seen with certain azimuths of the illumination ; the 

 bright dots when the light enters from the base of the scale ; the 

 fainter ones when it enters from the opposite extremity; and the cy- 

 lindrical appearances when it enters at the side of the scale, in the 

 longitudinal direction of which these appendages lie. 



" Again, with the scale of the Lepisma, this is clearly seen to be 



VOL. V. 2 I 



