About the satellites Mars possesses, Deimos and Phobos, we may perhaps say a word, as recent knowledge concerning them exemplifies the care now taken to such ascertainment and the importance of considering factors often overlooked. Soon after they were discovered in 1877, they were measured photometrically, with the result of giving a diameter of six miles to Deimos and one of seven miles to Phobos, and these values unchallenged entered the text-books. When the satellites came to be critically considered at Flagstaff, it was found that these determinations were markedly in error, Phobos being very much the larger of the two, the actual values reaching nearer ten miles for Deimos and thirty-six for Phobos.
In getting the Flagstaff values, the size to the eye of the satellite was corrected for the background upon which it shone; for the background is all-important to the brilliancy of a star. In the case of a small star near a planet, the swamping glare of the planet is something like the inverse cube of its distance away. Furthermore, the Flagstaff observations indicated how the previous error had crept in. For before correction for the differing brilliancies of the field of view, the apparent size of the satellites judged by conspicuousness was about six to seven. The photometric values must have been taken just as they came out, no correction apparently having been made for the background. Now the
