hearing upon the Classification of Spiders. 309 



has been suggested * that the anterior auxiliary mammillae 

 of Liphistiiis correspond morphologically to an unpaired 

 process called the coluhis, which is found between the ante- 

 rior mammillge of many spiders, e. g. Epeira. If the colulus 

 is not found in any spiders that possess the crihelliwi, it seems 

 to me probable that Dr. Thorcll's suggestion is correct. But 

 if the crihellum and colulus coexist in any spider, it is clear 

 that either my suggestion or Thorcll's is erroneous. 



If the homologies that I have suggested above are correct, 

 some of the Dipneumones at least possess representatives of 

 all the eight mammillaj of Liphisfius ; but this is not the 

 case with any of the Tcrritelarice. In this group the spinners 

 are nearly always arranged in two pairs — an anterior, con- 

 sisting of two short one-jointed segments, and a posterior, 

 consisting of two long three-jointed segments. How these 

 mammillae are to be correctly compared with those of Liphis- 

 tius or of the Dipneumones is to me by no means clear. The 

 posterior pair may be homologous to either of the principal 

 pairs of LipMstius and the anterior pair to either of the 

 auxiliary pairs of this animal ; or the two pairs may corre- 

 spond to the two principal pairs of Liphistius. But in either 

 case the disappearance of two pairs has to be accounted 

 for. Some of the Territelarise, however, such as Pelecodon 

 and Ilexathele, have six raammilla3, the additional ones being 

 short and placed in a transverse line with the ordinary ante- 

 rior pair. But the anterior series is not alike in the two 

 genera, the two internal mammillas being considerably larger 

 than the two external in Pelecodon, the converse obtaining in 

 Hexathele. This renders a comparison between them a matter 

 of some difficulty. I venture, however, to make the following- 

 suggestions on the point. In Pelecodon the large internal 

 pair is homologous to the anterior pair of the Dipneumones 

 and of the anterior principal pair in Liphistius, the smaller 

 external pair being the horaologues of the intermediate pair 

 of the Dipneumones and of the posterior auxiliary pair of 

 Liphistius. If this be so, the last-named mammilla3 have 

 shifted their position so as to lie completely in front of the 

 posterior mammillfe. As regards Hexathele, it seems reason- 

 able to suppose that the mammillae that are present arc the 

 same as those that are developed in Pelecodon. They may, 

 too, correspond exactly in position although differing in size. 

 An alternative hypothesis, however, is that the large pair of 

 this anterior series in Hexathele corresponds to the large ones 

 in Pelecodon. In this case the small intermediate pair in 



* Vide Thorell, Ann. Mus. Geiiov. xxviii. p. 29 (1889-90). 



