42 JOURNAL OF THE 



on either side, and pulled them nearly together, so that there 

 was just space enough left for it to see the ant when it crossed 

 the opening. When this happened, the spider threw the doors 

 wide open, caught the ant, and in the twinkling of an eye 

 had dropped back to the bottom of the tube with its game. 

 This I saw repeated several times during the months of Jan- 

 uary and P\^bruary. At one time I placed a large cricket in 

 the jar. The spider made several attempts to capture it but 

 was not strong enough. It would probably eat crickets 

 when fortunate enough to catch one. I think also that it 

 would eat small carabid beetles, for I found several in the 

 moss where I collected the spider. I have found the remains 

 of ants in the nests. 



I hav^e made no observations on Mf/DnrkiapIuJn/o/uf.'n in 

 captivity, but from the location of their nests it is quite 

 evident that the members of this species are extremely fond 

 ofants, and seek to build their nests either directly in an ants 

 nest or in close proximity to it. I think in some cases it is 

 intended that the main tube shall be used as a passage for 

 ants, so that by coming out of the branch the spider can 

 capture its food, for it is certain in these which I have ob- 

 served that the external door is either wanting or old and in 

 a neglected condition, while the door at the branch is kept 

 well repaired. If I am correct in this conclusion it would 

 account for the apparent variation noticed in the kind of 

 nest built by this species. In the nest represented in fig. i6, 

 Plate IV, the entrance b, and^the landing ((, a, of the ants' 

 nest answers for the main tube of the spider, and only the 

 one tube and trap-door is required for the use of the spider.' 



' Note. — There ^^viriently is quite a variation, even with the same 

 species, in the kind of a nt St which trap-door spiders mnlce. As these 

 prove to be very iDtelligent creatures, I beli« ve the vary in the «'ou- 

 siruction of ihe nest as ihe conditio s of their environment varies, an i 

 that tliey adopt that plan which provides for them the safest abode, and 

 at the sime time will serve them the best as a trap At the same time 

 we must not lose sight of a common typn, wliich, with some useful 

 variati(ms, each species follows; though specie- otdiffiijent genera may 

 n)ake the same kind of h nest, and be identical in food habit, srecies of 

 the same genus may yary with regird to the C')mmon type; so that »he 

 use of msts, and food habits, becomes of less \alue ihan has b eu h' re- 

 tofore supposed by S'>me f )r purp 'ses of clas-ification. 



Mr. Moggridge says, "H. A. and T. D, Spiders" Supplement, pp. 230, 

 and 237 ; the range and distribution of a sptcies largely I'epeud upon the 

 niture of its food, aud this will also bean indication of the rivals, etc, — 

 ''and in many cases even the .structure ^nd position of its dwelling place 

 will be governed by this same all — important question of food supply. 



Rev. O. P. Cambridge, at that time came to this conclusion, "and in 



