2. CALCULATION OF THE FACTORS FROM THE ABSORPTION OF THE PRISM. If the layer of fluor-spar of thickness / absorbs the fraction a, the emergent radiation from a prism of base B must be multiplied by the factor to obtain the incident radiation, as may be seen from a simple integration based on the law of absorption. The prism had its corners chipped, and had by no means a mathematically simple shape. Measurement of the base gave in the mean a value of In the following table & is the minimum deviation of the prism, the corresponding wave-length, and a the amount of the absorption I observed in a clear plate of fluorite of 4.056 mm thickness. The factor computed is calculated from the absorption, the factor observed is that determined from the energy curves above. Not until 9.324μ does the observed intensity have to be increased more than corresponds to the absorption of the prisms. This may be due to the absorption of the air of the room or to an erroneous determination of the absorption of the fluor-spar, but it has no bearing on our conclusions. The energy curves obtained with an ordinary lampblacked bolometer must lie too low even at short wave-lengths, representing a power of absorption of the lampblacked surface decreasing for increasing wavelengths. The layer of lampblack (through which the radiation has to pass twice) is therefore still considerably transparent even * In Wied. Ann., 53, 333, 1894, I give the observations of the total loss of light, from which 5.05 per cent. is to be deducted for loss by reflection. in a thick stratum, or else it has a rising reflecting power with increasing wave-length. From several experiments the latter seems to me to be the case. Any such deviation at these long wave-lengths would be naturally expected, since the absorption of the lampblack bolometer strip in the reflecting shell can only be approximately equal to unity. Although the arrangement of bolometers II to V was different and represented an increase of the blackening (passing from the imperfect to the perfect hemisphere), nevertheless they absorbed almost equally in this spectral region, so that they seem to justify the conclusion that they all approximate "the absolutely black body" in this region so nearly that the casual deviation cannot produce appreciable error. I therefore believe that it is permissible to employ the results of the measurements with these bolometers in the determination of the constants of law I. The constant c, is most accurately determined from these experiments by the fivefold value of the constant product m X T. This is in the mean: For bolometer II III IV V Mean of all. The single values of a series of observations which deviate most widely from the mean of all are: 2907 (Bolometer II, 450°, cavity with layer of copper oxide); 2884 (Bolometer III, 100.5°, cavity with layer of platinum-black). The first value is not entirely reliable, since it easily happened that some parts of the layer of copper oxide fell off, leaving the glass uncovered. This always gives occasion for an increase of the wave-length of the maximum. I estimate the highest possible error of the mean of all as something like 3 per thousand. Accordingly c2 will be = 5 X 2891 = 14455, with an error which at most I estimate at 40. Since in formula I, represents a number, c, must have the IK. Ångström has observed in thin layers an increasing transparency of lampblack with increasing wave-length. Wied. Ann., 36, 720, 1889. dimension: Wave-length × Temperature. Therefore C2=14455 × Degree Centigrade of absolute scale. The value of c, found by H. Wanner and myself' in a different region of temperature and wave-length by measurements of an entirely different kind was 14440, but it had considerably less accuracy. m The bolometric measurements at high temperatures recently published by O. Lummer and E. Pringsheim2 yielded values of X, XT from 2837 to 2928, mean 2879, for one series, and for another series, in which the exponent 5.2 held good in the general formula I, five values between 2766 and 2986, mean 2876. From these they deduce for values of c2, 14395 from the first series, and 14955 from the second series. Taking into account the errors possible according to the results of the observations and to the methods of reducing them, the agreement with my value seems to be good. F. Kurlbaum3 has measured the difference of the total radiation of a cavity coated with lampblack at 100° C. and at 0° in Watts x cm2, and has calculated that the heat equivalent of the total radiation emitted by a black body at the absolute temperature T amounts to The constant c1 has by formula I or formula II the dimen sions 1 Energy of radiation × (Wave-length)*. Their numerical result depends, therefore, on the unit of the This JOURNAL, 9, 304, 1899. Verhandlungen der Deutschen Physikalischen Gesellschaft, February 3, 1899. 3 Wied. Ann., 65, 746, 1898. 4 Loc. cit., p. 666, where for a 5, II (a2) 6. energy of radiation. In my experiments this was arbitrary and was different for the six different bolometers, so that a more precisely definite value of the quantity c, could not be computed from these experiments alone. In order to refer the measurements to a definite unit of radiation let us now determine the value of c1, with which the total radiation of my source of radiation reaches the quantity determined by Kurlbaum, so that the energy of radiation is measured in it follows that with the value c, 14455 X Temperature degrees = With these values of the constants the equivalent in work for the radiation of the black body of Temperature T, for any region of wave-length whatever, may therefore be computed by formula I. MINOR CONTRIBUTIONS AND NOTES THE THIRD CONFERENCE OF ASTRONOMERS AND ASTROPHYSICISTS. THE Committee charged with the selection of a time and place for holding a third conference of astronomers and astrophysicists met in the city of Washington on February 8, 1899, and by unanimous vote of the members present, Messrs. Newcomb, Morley, Hale, and Comstock, resolved that such a conference should be held at the Yerkes Observatory early in the following September, the precise date to be subsequently determined by Professor Hale. In accordance with this resolution, and at Professor Hale's invitation, the conference will be held at the Yerkes Observatory, Williams Bay, Wis., beginning on Wednesday, September 6, and closing on Friday, September 8. In its plan and scope this conference will be similar to those held in 1897 and 1898 at Williams Bay and Cambridge, accounts of which have been published in this JOURNAL and elsewhere. The committee charged with perfecting a plan for the organization of a permanent society of astronomers and astrophysicists to have charge of future conferences will present its report at this time. A circular giving information regarding local arrangements will be issued shortly. GEO. C. COMSTOCK, Secretary. |