Hastings — Error of Collimation in the Human Eye. 313 



temporal side in terms of tenths of the separation of the wires. 

 A subsequent determination of the absolute value of the 

 interval between the wires yields the real displacement. In 

 the observations which were made by Professor Beach, as well 

 as in those made by myself, the reduced interval was '075 cm , 

 so the probable error of our measures may be estimated at 

 about one-tenth of this quantity. 



A 



Fig. 1. 



Fig. 2. 



In the reduction of the observations it was assumed that the 

 radius of the cornea was, in every case, the same as that of the 

 schematic eye, namely, 0*7829 cm . Such an assumption was not 

 necessary, since it would have been easy to determine this ele- 

 ment for each case, but it was convenient and could lead to no 

 error which could become significant in any considerable group 

 of measures. The method of reduction will appear from refer- 

 ence to fig. 2. Here AC represents the axis of an eye with the 

 geometrical center of the cornea at C and its first nodal point 

 at 7i,. The direction n x a will be that of the source of light and 

 of the observing telescope. The image of the source of light 



Am. Jour. Sci. — Fourth Series, Vol. XIX, No. 112. 

 21 



-April, 1905. 



