in this way. More common, however, are unconscious changes in an originally quite veridic message. We easily see what we expect to see, but with great difficulty what we do not. This may be due to individual idiosyncrasy, or it may be due to a prevailing idea of the time, affecting people generally, in which we unwittingly share. Fashion is as potent here as elsewhere. The very same cause will show us at one time what we remain callously blind to at another. A few years ago it was the fashion not to see the canals of Mars, and nobody except Schiaparelli did. Now the fashion has begun to set the other way, and we are beginning to have presented suspiciously accurate fac-similes of Schiaparelli's observations.
In any observation, the observer is likely to be unconsciously affected in some way or other pro or con, which, from the fact that he is unconscious of it, he is unable to find out. The only sure test, therefore, is the seeing what no one else has seen, the discovery of new detail. Next to that is not too close an agreement with others. Inevitable errors of observation, to say nothing of times and seasons, distance and tilt, are certain to produce differences, of which one has ample proof in comparing his own drawings with one another. Even too close agreement with one's self is suspicious. In the matter of fine detail, absolute agreement is therefore neither to be expected nor to be desired.