The door is opened by pushing or pulling in any direction any one of the four strings which in Fig. 2 are shown in the niche to the right of the door. It is easily evident that the door may be opened in several different ways. It is thought that this seeming objection in actual experimentation proved a decided advantage with some birds.
All the birds are first fed by means of food placed in this box for some time before they are required to open the door. In making an experiment the food is placed on the inside just beyond the wires. The door is then closed. The birds are free at all times to move around the box. No bird has ever been confined in order that it might see another first open the door and then be liberated and allowed to try for itself. Such a method requires a break in the time which is certainly more difficult for the animal to cross mentally than it is for man.
As will be evident in many of the later series, the experimenter has sought whenever possible, by change to a kind of food more to the liking of some birds present in the experiment, by change in the positions of the food-pan, by changes in location of the strings on the box, by the introduction of strange, more pugnacious, younger, or more active and playful birds; by all these and other means the experimenter has sought to produce rivalry, competition, struggle, fear, new caution, interest and attention and above all a real necessity and opportunity for each animal to do something. When these conditions are obtained (and even then) our laboratory conditions are far enough from matching those in the birds' free life outside.
Some of the series may seem to the reader to have been carried to unnecessary length, yet such tests as laboratory studies permit are very foreign to birds and should, therefore, be given many repetitions. How foreign and artificial one can scarcely appreciate unless he has sat by the half hour outside the cage looking through the peep hole and watched the birds in the early tests doing even-thing but pulling or pushing the (to the experimenter) perfectly obvious string.
It is probably more natural for one bird to imitate another in song, or for a parrot to imitate the human voice, and yet how long often does one have to wait for results. Many repetitions of a word are often required before we expect to get any attempt at imitation. Artificial tests, then, can hardly err on the side of supplying too many opportunities for one bird to see another perform the act which is the model.
should be added, as well worth the time and patience required, that such changes as those mentioned above, particularly those of the location of the strings, are better calculated than almost any other to put to a real test the intelligence and
