This quiescence must be imitated by the observer, should he wish to watch the proceedings of the bird, as the least movement will startle it. The reason why so many persons fail to observe the habits of animals, and then disbelieve those who have been more successful, is that they have never mastered the key to all observation, _i.e._, refraining from the slightest motion. A movement of the hand or foot, or even a turn of the head is certain to give alarm; while many creatures are so wary that when watching them it is as well to droop the eyelids as much as possible, and not even to turn the eyes quickly, lest the reflection of the light from their surface should attract the attention of the watchful creature. One of the worst results of detection is that when any animal is startled it conveys the alarm to all others that happen to be within sight or hearing. It is evident that all animals of the same species have a language of their own which they perfectly understand, though it is not likely that an animal belonging to one species can understand the language of another. But there seems to be a sort of universal or _lingua-franca_ language which is common to all the animals, whether they be beasts or birds, and one of the best known phrases is the cry of alarm, which is understood by all alike. I need hardly say that it is almost absolutely necessary to be alone, as there is no object in two observers going together unless they can communicate with each other, and there is nothing which is so alarming to the beasts and birds as the sound of the human voice. Yet there is a mode by which two persons who have learned to act in concert with each other can manage to observe in company. It was shown to me by an old African hunter, when I was staying with him in the New Forest. In the forest, although even the snapping of a dry twig will give the alarm, neither bird nor beast seems to be disturbed by a whistle. We therefore drew up a code of whistles, and practised ourselves thoroughly in them. Then, we went as quietly as we could to the chosen spot, and sat down facing each other, so that no creature could pass behind one of us without being detected by the other. We were both dressed in dark grey, and took the precaution of sitting with our backs against a tree or a bank, or any object which could perform the double duty of giving us something to lean against, and of breaking the outlines of the human form. Our whistled code was as low as was possible consistent with being audible, and I do not think that during our many experiments we gave the alarm to a single creature. When the observer is remaining without movement, scarcely an animal will notice him. I remember that on one occasion my friend and I were sitting opposite each other, one on either side of a narrow forest path. The sun had set, but at that time of the year there is scarcely any real night, and objects could be easily seen in the half light.