61 



Journal," and for the full details of which we are indebted to the 

 courtesy of Prof. Keith. Nor is the interest of this elaborate 

 mathematical analysis appreciably lessened by Mr. Wenham's 

 doubt, as to the reliability of the data furnished by Mr. Tolles 

 as a basis for the computation ; since if the data do not accu- 



seenf to ' represent a practicable combination of lenses, which 

 might be made into an objective, and that is what we want to 

 know, and what the long discussion has derived its cat-like life 

 from. The objection in question consists of seven lenses ; a quad- 

 ruple back consisting of a double-convex of crown glass, a plano- 

 concave of flint, a plano-convex of crown and a meniscus of flint, 

 a double middle formed by the union of a double-convex of crown 

 and a double-concave of flint, and a simple hemispherical front of 

 crown. The plan of grinding the lenses as thin as possible is dis- 

 carded, as in much recent work, and some of the lenses are quite 

 thick at the thinnest point. The following figures represent the 

 data of construction, the letters a to g representing the seven lenses 

 in regular order, beginning with the upper lens of the back com- 

 bination (See also Fig. 24). 



Distance of back combination from middle -008. By screw 

 collar adjustment the front is set at a distance of '00528 from the 

 next surface. The light is assumed to start from a point ten 

 inches above the first (back) surface, and is traced through the 

 objective to a focal point below. The table on p. 62 represents 

 the distance from the axis at which the extreme ray crosses each 

 surface, and the angles which the ray, before crossing, makes with 



1 This gives a computed angular aperture of 110° 35' 10", which 

 largely exceeds the 87° obtained by measurement by Dr. Wood- 



