Januaey 16, 1903.] 



SCIENCE. 



83 



1889, when the telescope was dismounted. 

 I may say that, previous to this period, the 

 observations of phenomena have usually 

 been made by estimation. This was true 

 with regard to the determination of longi- 

 tude almost without exception, and very 

 few positions in latitude have ever been 

 determined with the micrometer. Ama- 

 teur observers, who have no driving clock 

 or micrometer, must necessarily rely on 

 eye estimates for longitude and latitude, 

 but when a telescope is equipped for 

 micrometer work there is no better excuse 

 for guessing than in the determination of 

 the distance of a pair of double stars. 



Sketches or drawings of the planet Ju- 

 piter are of very little value in the absence 

 of other data. It is not unusual to find the 

 latitude of conspicuous markings eight or 

 ten degrees in error, and longitude a cor- 

 responding amount. At the beginning of 

 my observations on Jupiter I decided to 

 fix the size and position of all objects seen 

 on the disk by micrometrical measurement. 

 By such a system of procedure positive 

 facts will be established, which in time 

 may enable us correctly to interpret the 

 complicated phenomena observed. 



During the past twenty-five years the 

 so-called canals and double canals on Mars 

 have been the subject of much discussion. 

 I believe if their position were fixed by 

 micrometrical measurements, we should 

 soon be able to decide what is real and 

 what is imaginary. 



In order to use the micrometer for 

 measurements on a planet, it is necessary 

 to know the size of the disk. Jupiter has 

 been measured by many astronomers, both 

 with the micrometer and with the heli- 

 ometer, but the measurements made differ 

 considerably, due to two causes. First, 

 irradiation, which depends on the size of 

 the telescope, or rather on the magnifying 

 power employed. Second, the increased 

 size of the image, due to the condition of 



the atmosphere. In the use of the heli- 

 ometer the true irradiation may be elim- 

 inated, but not the increased size of the 

 disk due to definition. In any case the 

 measured size of the disk depends directly 

 on the magnifying power employed. 



In 1880 I made a series of measures of 

 the polar and equatorial diameter of the 

 planet with powers of 390 and 638, and 

 in 1897 a series of measurements with 

 powers of 390 and 925. In all cases, 

 whatever the condition of the seeing, the 

 lower power gave the larger diameter. 

 From the measures made on six nights in 

 1897, when the seeing was good enough to 

 be able to use a power of 925, the differ- 

 ence for the two powers employed was: 

 polar, -f-0.27"; equatorial, +0.31". In 

 1880 for ordinary seeing the difference for 

 the two powers employed amounted to 1". 

 In order, therefore, to have some standard 

 of size it would be necessary to decide 

 upon the magnifying power employed with 

 which the measures were made. Because 

 of this apparent change in the size of the 

 disk due to definition, to locate with pre- 

 cision any object on the surface of the 

 disk, or a satellite off the disk, it is neces- 

 sary to refer the object to both limbs of 

 the planet at the time of observation. If 

 the object is referred to only one limb, 

 under unfavorable atmospheric conditions 

 an error of 1" of arc would be easily pos- 

 sible, but if it is referred to both limbs, 

 then the effect of the irradiation, or en- 

 largement of the disk, is almost wholly 

 eliminated. In the reduction of my mi- 

 crometrical work on Jupiter I have used 

 the values 18.33" and 19.48" for the semi- 

 axes of the planet at mean distance. 



These values for the size of the disk 

 were found from a great many differential 

 measures made in 1880-1 with a power 

 of 390, and are somewhat larger than 

 those given by the heliometer, owing to 



