crests (whether they are unbroken throughout or ragged in places) is not so 

 uniform. 



[With regard to the sculpture and armatiu'e of the antero-lateral borders 

 of the carapace, and with regard to the epigastric and post-orbital crests, I 

 should like to say that, however important they may be for discriminating 

 species, they do not of themselves alone give any clue to the affinities of species.] 



(8) The shape of the abdomen of the adult male is a safe guide to 

 affinities, and the shape of the two terminal segments, and sometimes of the 

 fifth segment also, is often of diagnostic value for species ; but this holds good 

 only for the full-grown adult. 



(9) Antennal flagdlum : this has not been used much for diiferentiating 

 species, but its taxonomic value is often very striking. 



(10) Mandibular palp. Acting upon a suggestion made by Dr W. T. 

 Caiman, I have examined the mandibular palp in almost every individual of 

 the collection described in the sequel. It never varies. Though of no value 

 for differentiating .species, it seems to me to be the long-desired index of 

 natural affinities of species. 



(11) External maxillipeds:. The exopodite usually has a .strong plumose 

 flagellum, but there is quite a number of species in which the exopodite is 

 non-flagellate. In some individuals of these non -flagellate species there may 

 be a vestige of a flagellum — a mere papilla ; and in one of these species some 

 individuals may have, and sometimes on one side only, a well-developed 

 flagellum. Some stress has been laid upon a longitudinal groove on the ventral 

 surface of the ischium ; but in species where it is distinct its position may be 

 inconstant, and in species where it is indistinct its degree of distinctness may 

 be variable to the vanishing-point. The shape and breadth of the merus are 

 very constant. 



(12) Chelipeds. The shape of the dactylus, which also determines the 

 gape of the fingers, varies, and in the case of the larger cheliped (of the male 

 in particular) the amount of difference, depending upon age, may be enormous. 

 The chelae of the larger cheliped seem to go on growing throughout life, and 

 the most common growth-change is for the dactylus to become more and more 

 curved and toothless. The inference is that distinctions based solely upon the 

 form of the chelae and the gape of the fingers are to be distrusted. 



(13) Legs. As the proportions of the legs do not alter much with age, 

 whereas in many species those of the chelipeds do, distinctions based solely upon 

 the length of the legs relative to the length of the chelipeds may be fallacious. 



In deciding the limits of species, I have estimated all these points to the 

 best of my ability. In dealing with large series, I have often had the greatest 

 difficulty in deciding between "species "and "variety": I can vouch that my 

 " varieties " have cost me much work of the kind performed by King Bruce's 

 spider. 



