626 LIFE : OUTLINES OF GENERAL BIOLOGY 



relative, this South African spider does not wait for the appearance 

 of a moth at which it may hurl its globule, as the horseman his 

 bolas; for what she does is to anticipate events by whirhng the 

 globule round and round in a horizontal plane, though there is no 

 booty in sight. It must be noted in regard to both observations that 

 spiders generally are very short-sighted. In any case, the fact i^ 

 that the African spider keeps wliirling her globule round, minute 

 after minute, even for a quarter of an hour without stopping. 

 Pausing then, if she has been so long unsuccessful in catching a 

 moth, she draws up the thread and swallows the viscid globule, for 

 it loses its stickiness after prolonged exposure to the air. After a 

 short rest she makes another globule and repeats the extraordinary 

 performance, which is sooner or later rewarded by catching booty. 

 Neither of these two spiders makes an insect-catching web — not at 

 any rate during the cocoon-making season, to which the observations 

 were restricted. 



Whatever the definitions may be, there seems no doubt as to 

 there being a real difference between instinctive and intelligent 

 behaviour. But it does not follow that an animal predominantly 

 instinctive cannot also act at times intelligently, or that in the 

 course of instinctive behaviour there may not be an interpolation 

 of intelligent control. How should we know if such interpolation 

 occurred ? 



As a rough-and-ready test among animals of "the httle-brain 

 type", such as ants, bees, and wasps, one may say, as already 

 explained, that instinctive behaviour is marked by its routine 

 character, by being performed with a considerable degree of per- 

 fection the very first time, by being shared equally by all members 

 of the species of the same sex, and by remarkable limitations when 

 something goes a little agee. Our question is whether we can find 

 convincing instances of the interpolation of another kind of be- 

 haviour, which we cannot describe without giving the animal credit 

 for some understanding of the situation, some perception of relations, 

 some perceptual inference. We submit the following observations by 

 Rau as a typical case of what we mean. One of the hunting wasj- 

 a species of Pompilus, is in the way of pursuing trapdoor spidti 

 It opens the silken-hinged lid of the shaft and descends; it stings 

 its victim and drags it above ground to its own nest — to serve in 

 a paralysed condition as part of the larder of fresh meat for tlit 

 wasp-grubs when they are hatched. So far there is nothing very 

 remarkable, as wasps go; but let us continue the story. When the 

 autumn comes, the trapdoor spider in question makes a second 

 shaft to her burrow, diverging at an angle, and opening by a door 

 on the surface not far from the original entrance. It is no longer 

 easy for the Pompilus wasp to catch its spider; for the burrowi 

 may escape by the second door, while the wasp is entering by the 



