2 Prof. T. Thovell on the 



features in their organization, the importance of which for tlie 

 purpose of systematization seems not to have been sufficiently 

 appreciated. 



The principal material on which this essay is based con- 

 sists of the collections in the National Museum at Stockholm 

 and the Gottenburg Museum of Natural History ; and I 

 avail myself of this opportunity to express my thankfulness 

 to the keepers of those institutions, Prof. C. Stal and Mr, A. W. 

 Malm, whose obligingness enabled me to study the scorpions 

 committed to their care. 



De Geer was, as is well known, the first who divided the 

 genus Scorjno, Linn., into smaller groups. He chose as the 

 basis for his classification the differences in the number of the 

 eyes. Most subsequent writers who have treated of the clas- 

 sification of scorpions [e.g. Leach, Hemprich and Ehrenberg, as 

 also C L. Koch) have either exclusively or principally adopted 

 the same principle of division , C. L. Koch's arrangement is still 

 employed by many naturalists, notwithstanding that Gervais 

 and, subsequently, Peters have clearly shown the compara- 

 tively trifling importance and the often variable and unsatis- 

 factory nature of the characteristics afforded by the eyes. 

 The merit of having first disengaged himself from the ordinary 

 view relative to the classification of scorpions belongs without 

 doubt to Gervais *, although even he appears to attribute too 

 much importance to the characteristics derived from the num- 

 ber of the eyes. The " groups," however, into Avhich he 

 collects the " subgenera " of his genus Scorpio are almost all 

 perfectly natural, and agree in part with those proposed by 

 Peters. His first group, containing the subgenera Andro- 

 ctonus, Centi-uncs, and Isometrus or Atreus, corresponds with 

 Peters's Androctonini and Centrurini (which I unite in one 

 under the denomination Androctonoidce) ; his second group, 

 TeiSgoneSjis identical with Peters's TeJegonini. The four fol- 

 lowing subgenera (ButhuSj Chactas, ScorpiuSj and Ischnurus) 

 he does not, however, like Peters, unite in one similar group, 

 but considers each of them as forming a separate group. 

 (The genera Vejovis and Dacurus = Centrums, C L. Koch, 

 which appear to have been unknown to him, were erroneously 

 placed in his first group.) 



But if Gervais's division is, on the whole, quite natural and 

 far better than C. L. Koch's, it nevertheless leaves much to 

 be desired as regards the sharp and sure limitation of the 

 groups ; and it is principally in this respect that the system 



* Gervais, " Remarques siir la fam. d. Scorpions/' in the Archives du 

 Mus. d'Hist. Nat. iv. pp. 201-240; Walckenaer et Gervais, Hist. Nat. d. 

 Ins. Apt. iii. pp. 32-74. 



