lieves that it is nearer 8".91. The first value corresponds to a distance of 92,380,000 miles, the second to one of 91,730,000. It follows that we have heretofore made an error of about three per cent, in estimating the distances, and about ten per cent, in estimating the masses of the solar system. Neither authority regards his result as more than approximative, Prof. Newcomb, for instance, considering that his own may, as likely as not, be over a hundred thousand miles from the truth.
We get no idea from these large-sounding numbers of the all but inconceivable minuteness of the error of observation which would cause them; and such a measure of uncertainty, far from casting any discredit on the exactness of modern astronomy, is an evidence of its surprising advance toward absolute truth. Modern astronomy began with the age of Kepler; but, while the angle which represents the error in the parallax Kepler found, would correspond to that filled by the width of one of the pages of this magazine at a distance of 2,000 feet from the eye, the error now admitted as probable by Prof. Newcomb is represented by a less angle than that filled at the same distance by the same leaf turned edgewise.
Now that we have considered the delicacy of the measurements which have already been made, we are prepared to appreciate the task of those who, on the 8th of this month, are about to try to better them, and to examine the principles underlying the methods which will be actually used in the trial. To do this, we may, perhaps, here recur to a former illustration. If we suppose a person looking at a remote object—let us say a lighted window—from a distance which is quite half a mile, the distance between his eyes bears nearly the same relation to that of the light, that the distance between any two stations practically usable on the earth does to that of the sun. Accordingly, the difficulty of obtaining the sun's parallax, without moving: from off the earth, is the same in degree that the observer would experience in measuring the distance of the light without moving from his place, and by means of the small virtual change of his point of view, obtained by looking at it with either eye; and it is under such all but insuperably hard conditions that astronomers will actually be working this month.
To see how Venus comes to their aid, we may represent her motion by a car moving at a uniform rate on a circular track, between the light and the observer. If the car pass across the light from left to right (as Venus crosses the sun), it will of course cut off the observer's view of the left side of the window from the left eye first, and, if the motion be slow enough, we may suppose him to note the exact time before the sight of the same point by the right eye is intercepted.
If he know from previous watching how long it takes the car to make its whole circuit of 360°, he knows from his watch, by an ob-
