[ 467 ] 



XLIII. On the Curvature Method of Teaching Optics. 

 By C. V. Drysdalb, D.Sc* 



NOTWITHSTANDING the fact that the wine theory of 

 light lias been employed to demonstrate some of the 

 more simple problems in the domain of what is gefieyally 

 termed geometrical optics, and with manifest simplicity and 

 convenience, this appears to have been done rather with the 

 object of veri tying the wave theory than of showing how the 

 subject of optics can be completely dealt with from this 

 standpoint : and few men of science or teachers of optics 

 appear to have realized the advantages of physical methods 

 both for practical work and teaching, and that they should 

 entirely supersede the geometrical or ray methods. This is 

 unquestionably due to a very large extent to unfamiliarity, 

 and to the cramping effect of our university curricula and 

 text-books ; but a possible factor in the question is the im- 

 pressioo which seems to be prevalent, even among optical 

 specialists, that the physical methods have to be abandoned 

 at a certain stage, and that the more complex problems 

 relating to lens systems and aberrations must be treated by 

 geometrical methods. Owing to the fact that British men 

 of science have been actively engaged in extending the wave 

 theory towards penetrating the more fascinating mysteries 

 of interference, polarization, and electromagnetic theory, the 

 practical applications of optics have passed for the last half 

 century into the hands of the Germans, who took from us 

 the geometrical methods then in voffue and have since ex- 

 tended them with such marked success as to give the 

 impression that these geometrical methods are the most 

 suitable for the purpose. At the present time a strong 

 attempt is being made to revive the study of technical optics 

 in this country, unfortunately with great difficulties, owing 

 to the lack of satisfactory teaching ; and the object of this 

 paper i> to show that not only are physical methods the most 

 suitable at the outset, but that they are capable of being 

 employed with the same increased simplicity in the whole 

 domain of "geometrical optics.''' 



A few words are desirable at the outset as to what has been 

 done in the application of physical problems to reflexion and 

 refraction. The first step in this direction appears to have 

 been made by Herschel in 1827 f, who seems to have dealt 

 with ordinary lens problems fairly completely and to have 

 devised a very satisfactory curvature notation. According 



* Communicated by the Physical Society : read February 24, 1905. 

 t Herschel, Encyc. Metropolitana, 1 827." 



