1873.] 



for a Microscope Object-glass. 



115 



the marginal rays through the combination, it will be seen that, though 

 the successive refractions are nearly equalized, the contact-surfaces of the 

 middle pair are somewhat deep, though no over-correction existed or 

 was needed here, for this would have required a shorter radius still (the 

 density of the flint in this was 3*686). If this pair of lenses were not 

 cemented with Canada balsam, total reflection would take place near the 

 circumference of the contact flint surface, cutting off the marginal rays 

 at a, and limiting the aperture. It might be argued that practically this 

 would be no disadvantage, as these surfaces are united with Canada 

 balsam, whose refraction is higher than the crown ; so that the rays in this 

 case must proceed with very little deviation. But incidences beyond the 

 angle of total reflection may be considered detrimental, as they imply 

 excessive depth of curvature ; this can be discovered by looking through 

 the front of an object-glass held close to the eye, any air-films in the 

 balsam near the edge of the lens appearing as opaque black spots. 



At the commencement of the present year the author caused a few 

 object-glasses to be made, with a middle of the form of fig. 5, the per- 

 formance of which was very satisfactory. In this the extreme rays pass 

 at more favourable incidences, and within the angle of total reflection. 

 The upper lens is of dense flint. 



When the experiments on the single front were concluded, and the re- 

 markable corrective power of the triple back in conjunction therewith 

 had been proved, the next attempt was to make the middle also a single 

 lens, leaving the entire colour-correction to be performed by the one bi- 

 concave flint in the back. After numerous trials it was found that though 

 something like over-correction or negative aberration could be obtained 

 with the back, in the degree requisite for balancing the under-correction 

 of the single middle and front when set at the prescribed distance of the 

 aplanatic focus, yet by trial on the mercury globule all the results in- 

 variably displayed two separated colour-rings : these could not be com- 

 bined by any alteration in the radius of the lenses. By projecting the 

 blue and red, or visible rays of greatest and least refrangibility through 

 the system, the cause became apparent. The left-hand section of this 

 object-glass is shown in fig. 6. The rays from the focus are slightly 

 divided by the first front surface. On .emerging from the back the sepa- 

 ration is increased ; the red ray (r) is outwards, and the more refran- 

 gible or blue ray (b) inwards. Next, the divergence of these two rays is 

 extended by the middle single lens. The following crown lens extends 

 the angle of divergence so far that the flint lens of the back triple cannot 

 recombine them ; and they emerge at two distinct zones, shown by the 

 practical test of the " artificial star " or light-spot reflected from a 

 mercury globule, viewed within and without the focus. 



It might be supposed that these rays at their final emergence can be 

 so refracted as to project the blue outwards. A crossing point would 

 then occur at a fixed conjugate focus in the body of the microscope, at 



