144 re;ports of investigations and projects. 



under observation could be simultaneously directed to the slit of a Littrow 

 grating spectrograph, of i8 feet focal length, and to the slit of a rotating- 

 mirror apparatus for measuring the frequency of the spark. 



The investigation was made with two principal objects in view : First, plates 

 have been taken of the titanium spark under various conditions of induc- 

 tance, capacity, potential, relative length of first and second spark, and with 

 and without air-blast, for the purpose of getting accurate knowledge of the 

 amount of variation which the lines undergo. Second, photographs of the 

 spectra of the sun-spot metals titanium, vanadium, iron, chromium, manga- 

 nese, nickel, cobalt, and magnesium, extending from the yellow to about 

 X 3600, have been made for the purpose of obtaining an accurate measure of 

 the relative enhancement of the lines of these metals. These plates contrast 

 the spectrum of the comparatively unimpeded condensed spark of high fre- 

 quency with that of the impeded spark of a frequency about 75 times slower, 

 the changed frequency being produced by a variation of inductance alone. 

 The intensities of the lines were measured with a Hartmann photometer, 

 in which a wedge cut from the plate to be measured, and formed by a 

 penumbral method, is employed. The curve of these wedges, which is a 

 straight line, is checked or calibrated by a standard intensity interval and 

 a number of intervals of varying intensity photographed on the same plate. 

 As the result of a preliminary measurement of some titanium spectra, it 

 appears that changes of any nature in the discharge circuit which increase 

 the instantaneous current at the moment of discharge increase the relative 

 intensities of the spark lines. If the inductance is varied, all other factors 

 remaining constant, it is found that the intensities of all lines hitherto investi- 

 gated vary approximately in direct proportion to the frequency of the spark. 

 The frequencies are accurately compared with one another by means of a 

 device which mechanically throws into the discharge circuit, for an instant, 

 a standard inductance. The accuracy with which the relative intensities of 

 the lines can be measured for any given condition of the spark, as determined 

 for a comparison of various photographs taken with different exposure times, 

 is about ± 2.5 per cent, when the light from various parts of the spark is 

 integrated by means of a ground-glass screen before the slit. The method of 

 measurement with the wide slit employed (0.008 inch) is convenient and 

 sufficiently accurate, since various settings on well-separated lines generally 

 agree to within less than 0.5 per cent. 



Pyrheliometric Observations. — The duties of the Smithsonian Astro- 

 physical Observatory in Washington prevented the continuation of the 

 Smithsonian Expedition's work on Mount Wilson during the summer of 

 1907. At Mr. Abbot's request, however, a pyrheliometer was used by Mr. 

 Olmsted on many days during the winter, spring, and summer, for the pur- 

 pose of determining the solar constant, on the assumption that the correction 

 derived from the Smithsonian Expedition's observations in 1905-6 can be 



