2l6 ENTOMOLOGISK TIDSKRIFT IQOI. 



the seventh — are adorned very difterently in some of the spe- 

 cies, and none of my species agrees completely in this respect 

 with another. An examination of several specimens of some of 

 the species gave the result that the individual variation between 

 adult specimens in the number of hairs was at most very slight, 

 and that long and especially stifif hairs and setae only in one case 

 (the group on the fourth segment in K. Diirabilis) varied in 

 number, and the difference was as slight as possible. Further- 

 more the three last segments present a specific character: in three 

 of the species these segments have one of their setœ in the 

 ventral median line, but in the three other species (A', mirabi- 

 lis, K. cJiilcnsis and K. Grassii) this seta is wanting. The 

 last segment in Ä. Grassii has only seven long setœ, in the 

 other species the number is a little higher and the greatest 

 number, ten setœ, is found in À'. angusta and K. siatnensis. 

 But these setse are often not easy to count with certainty. 



It may be inserted here that the middle portion of the side 

 of the second to sixth abdominal segments lacks hairs (pi. 2, 

 fig. 2 h; pi. 4, fig. 2 e). And I must rectify an error in fig. 3 

 (and the corresponding sentence on p. 232) in the paper of Han- 

 sen and Sörensen: the median one of the lateral hairs on the 

 second to the sixth segment shall be omitted. 



Before ending this chapter another thing may be mentioned. 

 The exact dorsal limitation of the two free thoracic segments is 

 better studied on animals treated with caustic potash than on 

 specimens in the normal condition. The material of my new 

 species is rather scarce, and I found it unnecessary to sacrifice 

 specimens for the study of the céphalothorax, as no difference of 

 any importance seems to exist. But fig. i a on pi. 4 elucidates 

 the feature that sometimes the thorax is reduced in length by 

 contraction to such a degree, that the median portion of the 

 wall of the head overlaps even the anterior part of the last tho- 

 racic segment, while fig. i b exhibits a specimen with distended 

 céphalothorax, in which both thoracic segments are quite free. 

 (The limits between the segments and the connecting membrane 

 were not visible on the object delineated). In the shape of the 

 head I find small differences between the specimens, but gene- 

 rally they are difficult to describe and of very little value, and 



24 



