8 HENRY A. ROWLAND 



of the length of light waves, and it was everywhere admitted that for 

 the most precise spectrum measurements they were indispensable. In 

 their construction, however, there were certain mechanical difficulties 

 which seemed for a time to be insuperable. There was no special 

 trouble in ruling lines as close together as need be ; indeed, Nobert, who 

 was long the most successful maker of ruled gratings, had succeeded in 

 putting as many as a hundred thousand in the space of a single inch. 

 The real difficulty was in the lack of uniformity of spacing, and on 

 uniformity depended the perfection and purity of the spectrum pro- 

 duced. Nobert jealously guarded his machine and method of ruling 

 gratings as a trade secret, a precaution hardly worth taking, for before 

 many years the best gratings in the world were made in the United 

 States. More than thirty years ago an amateur astronomer, in New 

 York City, a lawyer by profession, Lewis M. Rutherfurd, became inter- 

 ested in the subject and built a ruling engine of his own design. In 

 this machine the motion of the plate on which the lines were ruled 

 was produced at first by a somewhat complicated set of levers, for which 

 a carefully made screw was afterwards substituted. Aided by the skill 

 and patience of his mechanician, Chapman, Rutherfurd continued to 

 improve the construction of his machine until he was able to produce 

 gratings on glass and on speculum metal far superior to any made in 

 Europe. The best of them, however, were still faulty in respect to 

 uniformity of spacing, and it was impossible to cover a space exceeding 

 two or three square inches in a satisfactory manner. When Rowland 

 took up the problem, he saw, as, indeed, others had seen before him, 

 that the dominating element of a ruling machine was the screw by 

 means of which the plate or cutting tool was moved along. The ruled 

 grating would repeat all of the irregularities of this screw and would 

 be good or bad just as these were few or many. The problem was, 

 then, to make a screw which would be practically free from periodic 

 and other errors, and upon this problem a vast amount of thought and 

 experiment had already been expended. Rowland's solution of it was 

 characteristic of his genius; there were no easy advances through a 

 series of experiments in which success and failure mingled in varying 

 proportions ; " fire and fall back " was an order which he neither gave 

 nor obeyed, capture by storm being more to his mind. He was by 

 nature a mechanician of the highest type, and he was not long in devis- 

 ing a method for removing the irregularities of a screw, which aston- 

 ished everybody by its simplicity and by the all but absolute perfection 

 of its results. Indeed, the very first screw made by this process ranks 



