( 207 ) 



VII. — On a Caia-dioptric Immersion Illuminator. 

 By J. Wake Stephenson, F.E.A.S., F.E.M.S. 



{Bead 11th February, 1885.) 



The continued and remarkable progress wliicli has been made in 

 the development of the homogeneous immersion principle has 

 induced me to direct my attention to an immersion illuminator 

 which shall exceed in numerical aperture that of any objective 

 which has been hitherto constructed. Mr. T. Powell has now 

 produced an objective of N.A, 1 • 5 (that is with an aperture half as 

 large again as that of a dry objective of 180^), and a condenser in 

 excess of this appears indispensable. 



It is true that the full aperture of an objective can be utilized 

 by the vertical illuminator, but, as the illuminating beam is on a dry 

 object reflected from the under side of the diatom, it necessarily 

 follows that such beam passes twice through the object (acting in a 

 manner somewhat analogous to Lecount's polariscope *) ; but if 

 dififraction spectra are generated on the side in contact with the 

 cover-glass, they must subsequently pass through the same struc- 

 ture from which they originated, and evidently in some cases 

 produce confusion of image. 



It is unnecessary to dwell further on this point, as we all know 

 that the appearances presented by Coscinodiscus and some other 

 diatoms, when illuminated in the ordinary way, give no indication 

 of the structure of the two individual plates of which those diatoms 

 consist, the passage of the light through the lower side of the 

 valve producing confusion when passing through the upper. 



The cata-dioptric illuminator which I now describe has an 

 aperture which is only limited by the refractive index of the flint 

 glass used. The aperture of the one on the table ranges from 

 to 1 • 644 N, A. on a flint-glass slide of suitable thickness, and 

 from to 1"512 on a crown-glass slide of the same thickness, 

 the glass having a refractive index of 1 '52. 



The first question which naturally arises is, what is the use of 

 an aperture of 1 • 644 N.A., when no objective which has hitherto 

 been made has an aperture exceeding 1 • 5, as far at least as I am 

 aware ? 



The answer is that with an illuminator having a considerable 

 excess of aperture, it is much easier to fimd what is required, than 

 it is with one which by calculation gives us exactly what we want 



_ * In this instniment the polarized beam passes downwards through the 

 object, is reflected from a plane mirror beneath, and again passes through the 

 object before reaching the analyser — so that the colour produced by a film of 

 selenite indicates a plate of twice the thickness of that which is actually under 

 observation. 



