the Grebe returned to the nest; but, on the Moor-hen's again descending, he again swam towards it, this time in a little threatening rush, driving it right up the bank. This took place yet a third time, and then, before descending the bank again, the Moor-hen walked some way farther off along the top of it. This, it must be remembered, was not under the actual stress of an attack, but deliberately, and though it was evident that the particular spot off the water which the Moor-hen was thus leaving was the particular spot where it wished to be. Two attempts to return and a previous lengthy occupation are sufficient to show this. It is in little everyday things like this, I think, that one can best trace the working of reason in animals. Elaborate experiments in which they are placed under quite artificial conditions are of little value. I have read a whole series of such where Cats and Dogs were put into boxes which opened by a certain mechanical contrivance, and it seemed to be expected of them that they should calmly examine the interior with a view to piercing into this, the result, of course, being held to show that they had no reasoning power. Probably as long as they were there, their mental distress and confusion was such that they had not. When Foxes, however, and Wolverines walk round traps and examine them (as to which see Professor Romanes' 'Animal Intelligence') they are calm, and have their wits about them. Moreover, they have had time and that kind of vivid experience which impresses things on the mind—factors which, in the wooing of reason, are found sometimes to be almost as much needed by men as by animals. There is too much tendency, I think, to go by experiments made in the study rather than by those which nature may be said to make. The reason of this is not difficult to understand. Men, as a rule, are more comfortable in their study, and they admire their own or each other's ingenuity. But the greatest ingenuity can hardly ever give what is the most absolutely essential factor in all experiments where animal psychology is concerned, viz. natural—or at any rate accustomed—conditions. I therefore think that to watch an experiment made by nature is in nine cases out of ten much better than to make one oneself.
The Moor-hen seemed to know that the Grebe would not follow him up the bank, for always, though he might be a little
