KEPORT OF THE SECRETARY. 



91 



receive from the center' of the solar disk as unity, then as we measure farther and 

 farther from the center the intensity will diminish, owing to the absorption the rays 

 experience in the ever-increasilig length of path they traverse in the solar envelope. 

 Different wave lengths are differentlv diminished. The figure gives the fractions 

 remaining for the intensity at points distant 95 per cent and 98 per cent of the radius 

 from the center. Referring to Fig. 3, we see that in both the solar and the terrestrial 

 atmosphere the absorption is least in the infra red, and indeed follows a somewhat 

 similar curve throughout the range of wave lengths here investigated. The accuracy 

 of the results is not great enough to insist on the significanc* of the inflections which 

 appear in the curves given, but here also there is a similarity between the terrestrial 

 and solar curves. Owing to the imperfection of the large solar image, due to "l)oil- 

 ing" of the air, to the rotation of the field ]iy the siderostat, and to the fact that the 

 image was enlarged l)y a convex mirror from about 3 centimeters diameter to 40 

 centimeters, it seemed useless to attempt iirojected further measurements, including 

 some on the radiation of sun spots, until some better way of producing the solar 

 image was arranged for. Experiments which have been made for this purjiose will 

 be descril)ed under another ca})ti(in. 



Fi{,'. 4. — The Inin.sparency of tlie solar envelope. Horizontal scale represents wave lengtlis. Vertical 

 scale exhibits relative amount of radiation from the edge and center of the sun. Curve 1 is for 

 points 95 ])er cent of radius distant from smi's center. Curve 2 is tor points 98 per cent of radius 

 distant from sun's center. 



I take mucli pleasure in sajdng that the work of taking energy curves foi' deter- 

 mining solar and terrestrial absorption, the measurements <if oi'dinates and areas, 

 and preliminary reductions, onlj' a very few of which 1 have here given, have bccti 

 etiiciently done by the junior assistant, Mr. F. I']. Fowle. 



iSt'iidtipc (jdlniiiiintrtir. — In my last report I stated at s<iine leiiglli tlie |)rogr(.'ss made 

 and looked for in increasing tlu^ working sensitiveness of the galNaiiomctcr. it was 

 stated that a modified form of needle system was projiost'd, mucli iieavier and 

 probably not less sensitive than those then in use. Tiie (k'sign there referred to was 

 based j^artly on experiments which had been made wliich showed that two f^mall 

 thin magnets could be placed within a distance ecpial to their own dianaeter without 

 suffering nmch loss of combined magnetic moment. If this proved ai>plicable to a 

 system of numerous magnets it would be possible to increase the nund)er of mag- 

 nets to a hundred or more without loss of sensil)ility overbalancing the gain which 

 would result from decreasing the relative moment of inertia of tlie nouuiagni'tic 

 material. Shortly after my return from Sumatra a needle system of this kind was 

 constructed, containing 120 magnets, 60 at each end, separated b}^ spaces equal to 

 the thickness of the needles. The problem of arranging them thus proved a pretty 

 difficult one, but w^s solved, though after a fashion not quite satisfactory. It was 

 a disappointment to find, however, that the sensitiveness of this sy.stem was only 



