Preliminary Table of Wave-lengths of Enhanced Lines. 455 



manner described by Rowland, the camera and grating being at oppo- 

 site ends of an iron girder adjusted exactly to the radius of curvature 

 of the grating. It has been found that by carefully making the adjust- 

 ment the scale can be maintained constant within a very small limit, 

 and the conditions were such that it was possible to use plates 18 inches 

 long without bending. 



In some cases a Steinheil prism spectroscope was used. The disper- 

 sion arrangement of this instrument consists of four dense flint prisms, 

 three of 45° angle and one of 60°. These are fed by a collimator of 

 inches aperture and 18 J inches focal length. The camera objective 

 is a single quartz lens of 2 inches aperture and about 54 inches focal 

 length (for A 4340), the non-achromatism of which necessitates a con- 

 siderable inclination of the photographic plate to the axis of the lens. 

 The total deviation for the blue region of the spectrum is about 150°. 

 The scale of the spectrum is as follows : — 



D — F = 2| inches. 

 F-K = 6| „ 



The spark conditions were as follows : — The r Spottiswoode coil is 

 capable of giving a spark 42 inches long in air. For spectroscopic 

 purposes, however, a condenser is inserted in parallel with the second- 

 ary circuit, the length of spark then obtainable depending on the 

 capacity of the particular condenser used. These have varied from a 

 single gallon jar to a battery of twelve jars of about 15 gallons each, 

 and finally a plate condenser has been used, at the suggestion of Pro- 

 fessor Boys. 



This consists of thirty sheets of plate glass, 30 inches by 25 inches, 

 with tinfoil, 24 inches by 12 inches, between each pair. The spark 

 under these conditions varies from about 25 to 2 mm. in length, and 

 this was again further controlled in intensity and duration by a 

 secondary spark gap in series with the one containing the metallic 

 poles. 



The primary was fed from the street circuit at 100 volts, the usual 

 current employed being about 25 amperes. Interruption of the cur- 

 rent in the primary was by means of a mercury break actuated by 

 hand. 



In the case of the arc, the exposures lasted generally for about 

 fifteen minutes, while an hour and a quarter was the average time 

 given for the spark. 



My thanks are due to Mr. C. P. Butler, who was employed in taking 

 the photographs, to Mr. F. E. Baxandall, who is responsible for their 

 discussion, and to Dr. Lockyer for assistance in the preparation of this 

 note. The enlargement of portions of the arc and spark spectra of 

 titanium was made by Sapper J. P. Wilkie, R.E. 



