55 



for the Garijpidae Hans. "Tho femora of the 2 anterior pair of legs divided into 

 a longer pars basali.s and a shorter pars tibialis. All tarsi 2-jointed. 4 eyes" (49. 

 pp. 231—232). The first of these characters nuisl necessarily be at least modified 

 in the Garijpidae Hans, on acconnl of the slrncliire of the femora I — II in Garypi- 

 niis Dad., naturally belonging to this family (cf. below); for this genus has the pars 

 basalis and pars tibialis of almost equal length and the definition must according 

 to ElJingsen (1904. 70. p. 2) be altered in the following manner "The femora of 

 the two anterior pairs of legs divided into a longer pars basalis and a shorter pars 

 tibialis, or with a trocliantin". If it is quite correct to name the basal portion of 

 femur I in Garypiniis Dad. a "trocliantin" will be discussed later on (cf. below). 

 Hansen divided his Cheliferidae Hans, into two subfamilies the ChcUferinae Sim. 

 and the Cbiridiinae Hans., of which the former has a trocliantin on the first pair 

 of femurs, while the latter has none (49. p. 232). 



This wrong interpretation of the legs have also left other traces on Balzan's 

 system, because he was consetjuently unable to realize the great differences, found 

 in the number of the tarsi in the leg I— II of a Clithoniiis C. K. and an Ohisiiim 

 C. K. He was consequently bound to use the structure and the presence of the 

 galea as his main character. His first family the Pseudobisiidae Balz, is, if we omit 

 characters of minor importance, distinguished in the following manner (12. p. 539) 



"Chelae ante apicem super digitum mobilem processu pellucido subuliformi 



vcl i'nrcilliformi praeditae. Tarsi heteromeri hi- vel triarticulati". He divided 

 this family into two subfamilies, distinguished by minor differences in the structure 

 of the galea, viz. the Pseiidobisiinae Balz. (p. 539), and the Microcreagrinae Balz.; 

 about the value of these two groups, which were abolished by Hansen (49. p. 230) 

 and certainly ought to be so, I refer to the below as well as to Ellingsen (1901. 

 58. p. 87). His Obisiidae Tom. are characterized by "Chelae . . . ante apicem super 

 digitum mobilem processu pellucido baud armatae .... Tarsi heteromeri, bi- vel 

 triarticulati, vel cuncti biarticulati" (12. p. 545). This family was divided into two 

 subfamilies viz. the Obisiinae Balz. (Dad.) and the Chtboniinae Balz. (Dad.), of which 

 the former ought to be characterized by the tarsi, being two- or three-jointed, 

 while the latter is marked out by the large chelae, the calyciformed tibia of the 

 palps, and the tarsi, which are always two-jointed (12. p. 545). His third family, 

 the Tridenchthoniidae Balz., is characterized by the structure of the galea (12. p. 509). 



As the galea is an organ subject to great variations and even wanting in the 

 male of Ideoroncus sp. at least according to Banks (1895. 50. p. 11), Balzan is cer- 

 tainly wrong in using it at the principal feature in the definition of his families, 

 especially when we have such an excellent character in the number of the tarsal 

 joints of the two first pair of legs. Hansen consequently divides (49. p. 232) the 

 Hemictenodactyli into the Obisiidae Hans., characterized by "The tarsus of 2 fore- 

 most pairs of legs 2-jointcd" and corresponding to Balzan's Pseudobisiidae Balz, and 

 Obisiinae Bad., and the C/)//io/i/((/oe Hans., characterized by "The tarsus of two fore- 

 most pairs of legs 1-jointed" and corresponding to Balzans Cblboniinae Dad. and 



