The illustration which I have already used will serve excellently to show the general principles on which the value of a transit of Venus depends; and as, for some inscrutable reasons, any statement in which Venus, the sun, and the earth are introduced, seems by many to be regarded as, of its very nature, too perplexing for anyone but the astronomer even to attempt to understand, my talk in the next few paragraphs shall be about a dove, a dovecot, and a window, whereby, perhaps, some may be tempted to master the essential points of the astronomical question who would be driven out of hearing if I spoke about planets and orbits, ascending nodes and descending nodes, ingress and egress, and contacts internal and external. Suppose D, fig. 42, to be a dove flying between the window A B and the dovecot C _c_, and let us suppose that a person looking at the dove just over the bar A sees her apparently cross the cot at the level _a_, at the foot of one row of openings, while another person looking at the dove just over the bar B sees her cross the cot apparently at the level _b_, at the foot of the row of openings next above the row _a_. Now suppose that the observer does not know the distance or size of the cot, but that he does know in some way that the dove flies _just_ midway between the window and the cot; then it is perfectly clear that the distance _a b_ between the two rows of openings is exactly the same as the distance A B between the two window-bars; so that our observers need only measure A B with a foot-rule to know the scale on which the dovecot is made. If A B is one foot, for instance, then _a b_ is also one foot; and if the dovecot has three equal divisions, as shown at the side, then C _c_ is exactly one yard in height. [Illustration: Fig. 42.] Thus we have here a case where two observers, without leaving their window, can tell the size of a distant object. And it is quite clear that wherever the dove may pass between the window and the house, the observers will be equally able to determine the size of the cot, if only they know the relative distances of the dove and dovecot. [Illustration: Fig. 43.] [Illustration: Fig. 44.] [Illustration: Fig. 45.] Thus, if D _a_ is twice as great as D A, as in fig. 43, then _a b_ is twice as great as A B, the length which the observers know; and if D _a_ is only equal to half D A, as in fig. 44, then _a b_ is only equal to half the known length A B. In every possible case the length of _a b_ is known. Take one other case in which the proportion is not quite so simple:--Suppose that D _a_ is greater than D A in the proportion of 18 to 7, as in fig. 45; then _b a_ is greater than A B in the same proportion; so that, for instance, if A B is a length of 7 inches, _b a_ is a length of 18 inches.