It will be seen that, except for the July value, the size of the polar diameter comes out essentially the same throughout. Now, during July the polar cap was very large, and covered the southern part of the disk at the point where the polar diameter was measured. As it was much brighter than the rest of the disk, its irradiation must have been correspondingly great, and this would have had the effect of increasing the apparent length of the polar diameter beyond its true value.
The equatorial measures, on the other hand, show a systematic increase as the phase increased; and they do this on both sides of opposition. The increase, it will be noticed, is much greater than the probable errors of observation.
As the statement has been widely circulated that recent spectroscopic observations negative an atmosphere on Mars, it may be well to mention in a note that the observations in question neither affirm nor deny its presence, as their self-disclosed measure of precision, ¼ of an atmosphere, proves them incapable of it. They simply concur in showing that atmosphere to be thin. As a matter of fact, if spectroscopic observations did deny the existence of an atmosphere on Mars, such assertion would be fatal, not to the atmosphere, but to the observer or his instrument, as the existence of an atmosphere is demonstrated by the fundamental laws of