THE TRAP-DOOR SPIDER. 53 



actually in the face of a bank so as to be the cause of the trap-door 

 shutting to by its own weight ; but always each nest on a little bit of 

 flat, or almost flat ground." 



My observations differ from those recorded here in several 

 points. Near Lyttelton the nests are to be found on banks facing 

 all points of the compass, and in shady as well as sunny spots. Some 

 of the largest I have seen were in a small, moist, open-air fernery, 

 overhung by trees, and facing eastward. Again all the ground 

 around Lyttelton is more or less stony or rocky, and nests are 

 frequently found in places where there are only a few inches of soil 

 above the solid bed rock. They are frequently also to be found in 

 the face of a bank ; and more often than not their position is such 

 that the force of gravity must assist the door in shutting. In fact it 

 seems to me doubtful, whether the spider does ever instinctively 

 choose or make a flat surface for the commencement of its nest. 



The trap-door is almost invariably concealed, so that it can only 

 be detected by a close observer. With practice however the majority 

 can be recognised by their external appearance. In many cases the 

 mud on the door forms a protuberance slightly higher than the 

 surrounding surface. In one case I noticed as much as half an inch 

 of soil upon the lid. This is generally of considerable assistance in 

 enabling one to discover the nest in the naked soil ; but it often 

 occurs that one may suspect the presence of a nest from external 

 indications, but cannot be certain of it until the door ims been raised 

 with a stick or straw. In ground covered with vegetation their 

 discovery is rendered more difficult by plant growths which com- 

 pletely conceal the trap-door ; but I cannot think with Mr. Gillies 

 that the spider actually plants and cultivates these for the purposes 

 of deception. One of the best concealed doors that I have seen was 

 covered with moss, and so indistinguishable was it, that it could only 

 have been discovered by accident ; but of course there was a bed of 

 moss covering the ground for a foot or two all round it. 



With one exception all the doors, that I have found, have fitted 

 over the mouth of the nest, and not into it like a cork or plug. 

 In the exceptional case the hole was also peculiar, as it was the only 

 one that I have seen, which sloped upward from its entrance. The 

 majority, (particularly those on the sides of baiiks), make an acute 

 angle with the surface on their lower sides, and after continuing in 

 the same direction for two or three inches, bend nearly vertically 

 downwards. None of any length appear to be straight.. In the 

 exceptional case already referred to, the hole was aftewards found to 

 be only three inches long, and proportionately narrow. I removed 

 the lid and found that on the following day it had been replaced by 

 a similar one (i.e. one of the cork type). Unfortunately I did not 

 obtain the spider itself. With regard to the renewal of the lid, I 

 may say that I have observed it, in several cases. The work is 

 generally done in a single night, occasionally in two. Once I observed 

 the replacement in the course of a night of more than half a square 

 inch of surface, that had been sheared from the side of a nest with a 

 spade. 



My observations with regard to the holes themselves agree very 

 largely with those of Mr. Gillies. I found them to be from ten 



