2o8 KX lOMdI.OC.lSK riliSRKll' I' IC)OI. 



distance between the sues of the same j^air is a little shorter 

 than the length of each of them. 



In the other four species described in this paper 1 have 

 searched in vain for such sacs, and the ciuality of my material 

 enables me to deny their existence with certainty. But this cir- 

 cumstance renders it still more difficult to interpret the morpho- 

 logical significance or the function of the three pairs of sacs in 

 the two species. When Prof. A\'hkelp:r writes: »These sacs are 

 in all probability lung-books >>, 1 do not believe this interpreta- 

 tion to be a very probable one, as it seems to me very unlik- 

 ely that in some species such respiratory organs should be pre- 

 sent, while they are completely wanting in other species of si- 

 milar volume, and the species with sacs do not offer any other 

 character which could justify their reference to another genus. 

 As to the function of the sacs I am unable to set forth any 

 hypothesis tolerably probable to myself, and therefore I prefer 

 to give up the (luestion, leaving it undisturbed by airy specula- 

 tion to real study in nature. 



e. Flagellum. The material examined by W. Sorensen 

 and myself was, as mentioned on j). igç, very defective as to the 

 flagellum. Now I have examined a well-preserved flagellum of 

 K. inirabilis, K. WJicilcri and K. cJiilcnsis, besides a fragment 

 in two other species and a drawing of a third (A' Grassii). It 

 may perhaps be convenient first to describe the flagellum in K. 

 mirabilis and then to mention the differences observed in the 

 other species. 



The flagellum in A', mirabilis consists of thirteen oblong, 

 well developed joints, each with a whorl of very long, faintly 

 pubescent setae inserted on (or just beyond) the thickest part of 

 the joint, which is situated beyond the middle or at a rather 

 short distance from the end of the joint; on the last joint the 

 whorl is found a little beyond the middle, and near to the end 

 is seen another whorl of only four seta;, which are long, but yet 

 somewhat shorter than those in the first whorl and of similar 

 structure. The second, third, fifth, seventh and ninth joints are 

 conspicuously shorter than the others; these five shorter joints 

 and the l)asal one present besides a subapical whorl of very 

 consi)icuous glabrous setce, which are several times shorter but 



i6 



