The polar spots change considerably in size, as they do also in form. Sometimes they occupy nearly one-third of the disk, as is proved by many of my observations; while at other times they are so much reduced as to be totally invisible. It is to be remarked that the reduction of these spots generally corresponds with the summer seasons, and their enlargement with the winter seasons of the hemispheres to which they respectively belong. From these well-observed facts it would appear that a relation exists between the temperature of the two hemispheres of Mars and the variations of the white spots observed at its poles. A similar relation is known to exist on our globe between the progress of the seasons and the melting away and the accumulation of snow in the polar regions. Astronomers have been led, accordingly, to attribute the polar spots of Mars, with all their variations, to the alternate accumulation and melting of snows. On this account, the polar spots of Mars are sometimes also called "snow-spots." Errors have certainly been made by astronomers in some of their observations of the so-called polar snow-spots, other objects occupying their place having been mistaken for them. A regular series of observations on this planet, which I have now continued for seven years, has revealed the fact that during the winter seasons of the southern hemisphere of Mars, the polar spots are most of the time invisible, being covered over by white, opaque, cloud-like forms, strongly reflecting light. In 1877, during more than a month, I, myself, mistook for the polar spots such a canopy of clouds, which covered at least one-fifth of the surface of the whole disk. I only became aware of my error when the opaque cloud, beginning to dissolve at the approach of the Martial summer, allowed the real polar spot to be seen through its vapors, as through a mist at first, and afterwards with great distinctness. In this particular case, the snow-spot was considerably smaller than the cloudy cap which covered it, and it is to be remarked that it was not situated at the centre of this cloudy cap, but was east of that centre; a fact which may account for the so-called polar spots not being always observed on exactly opposite sides of the disk. From my observations of 1877, 1878 and 1880, it appears that at the approach of the autumnal equinox of the southern hemisphere of Mars, large, opaque masses, like cumulus clouds in form, began to gather in the polar regions of that hemisphere, and continued through autumn and winter, dissolving only at the approach of spring. These clouds, which varied in form and extent, were very unsteady at first, but as the winter drew nearer they enlarged and became more permanent, covering large surfaces for months at a time.