INSTINCT AS TAXONOMIC CHARACTER IN SPIDERS. 429 



distinctive is this habit in some families that on the strength of it 

 the great French arachnologist Simon placed the genus Rhoicinus 

 among the Lycosidae, although neither the disposition of the 

 eyes, nor other important morphological structures are of the 

 type to warrant the inclusion of the former in the very character- 

 istic family of ground spiders. Says Simon in his "Histoire 

 Naturelle des Araignees " : " Les quelques especes, pour lesquelles 

 j'ai propose le genre Rhoicinus, sont tres anormales pour la 

 famille des Lycosides, dont elles n'ont pas la disposition oculaire; 

 si Ton ne tenait compte que de ce caractere, on les rapprocherait 

 des Cybaeinx, particulierement des Campos tichomma, mais leurs 

 trochanters sont entailles d'une profonde echancrure apicale, 

 leur griffe impaire ne porte q'une seule petite dent basale et, de 

 plus, j'ai surpris Vune des especes portant aux filieres son cocon 

 globuleux, comme Vaurait fait un Lycosa." (Italics are mine, 

 A. P.) Without discussing the case in question since our know- 

 ledge of Rhoicinus is as yet very imperfect and I have no first- 

 hand acquaintance with the genus, I merely wish to point out 

 that of the characters mentioned by Simon the notch in the 

 trochanters is a character occurring in other families besides the 

 Lycosidse, such as Pisauridae, Clubionidae, Argiopidae, etc., the 

 number of teeth in claws is very variable, changing in other 

 families from genus to genus, while the disposition of the eyes is 

 undoubtedly a very old character and remarkably persistent 

 within families, being at the same time very little affected by age. 

 Thus the deciding character in this case remains the cocoon- 

 carrying habit and the question arises, which character is more 

 subject to modification, a fundamental, old morphological 

 character or a certain habit? 



On one of my excursions in Porto Rico I was fortunate enough 

 to make an observation which proves definitely that at least the 

 cocoon-carrying habit may be acquired by spiders the morpho- 

 logical characters of which prevent their inclusion in the family 

 Lycosidae beyond any possibility of dispute. While turning over 

 a rock in a field near Aguas Buenas I noticed two little spiders 

 on the underside of the rock, carrying their cocoons Lycosa- 

 fashion, attached to the spinnerets. Both proved to be new 

 species to which I have given respectively the names of Bathy- 



