ROTATIIIUA 169 



Tlie occurrence (if /\'. quadrata in the Nilyiris is of interest for the species is probably 

 absent in most truly tropical waters, occurring only within the tropics in elevated localities 

 under temperate cimditions such as are afforded iiy the aquatic iiabitats arnmid Ootacamund. 



Kcratella valga (Ehrenberg). 



In true K. valga the cycle (Klausener 1908, Hartmann 1918) appears to involve pri- 

 marily the addition of spines. The initial phase in the most complete cases is the form 

 called by Klausener K. curviconiis f. brchiiii. This form is figured as being somewhat longer 

 than K. quadrata f. cun-'icornis. In certain irregularities in the line of the posterior margin 

 of this figure there is, moreover, perhaps a hint of the position of the spine bases, demark- 

 ating a typical valga posterior margin, though this may be accidental. Fadeev (1927) has 

 described and figured as f. aspina a comparable form which is certainly clearly referable 

 to valga on the shape of the posterior margin alone without having to take into account the 

 other members of the cycle. An essentially similar elongate form (Figure 4 a) was recorded 

 from South Africa by Hutchinson and by Hutchinson, Pickford and Schuurman ( 1932) 

 as curvicornis. Klausener distinguished f. brehtiii from curvicornis by the fact that the 

 antero-median spines do not diverge in the former. This character is probably too vari- 

 able and in general does not separate valga from quadrata. It must be admitted that the 

 two species must often Ije hard to separate in their most reduced forms. 



In Hartmann's studies of this species the most exuberant form was one in which the 

 two posterior spines are subequal. This form seems to be the one figured by Schmarda 

 ( 1850, Plate IV, fig. Ill) as A. longicornis but it has doubtless been regarded by other authors 

 as actileata (i.e. quadrata s. s.). 



Apstein has described another exuberant form from Ceylon as tropica. This form is 

 characterised by the very long and unequal posterior spines, the left being about two-fifths 

 the length of the lorica without the anterior spines, the right about six-sevenths of this 

 length. This form, in spite of Hartmann's statement that it agrees well with liis form of 

 May, 1915 (f. 7'alga), differs from his figure of the latter in which the right spine is liardly 

 more than one-half the length of tlie lorica. It is clear from the work of Tschugunoff 

 (1922) and Fadeev (1927) that forms essentially similar to tropica are common in South 

 Russia, and Skorikov as long ago as 1896 figured (T. VIII, fig. 29) such a form from 

 near Kharkov. Jaku1:iski (1915) has figured a monospinous tropica form from Poland and 

 in South Africa, Hutchinson (1930) and Hutchinson, Pickford and Schuurman (1932) 

 fcnind tropica to be the only common form of ixilga. 



In all probability the most developed form of I'alga varies from place to place, and if 

 f. valga and f. longicornis represent the high;st development in some European localities, 

 tropica probably does the same in Ceylon, S. Russia, South Africa and Kashmir. The names 

 proposed for forms of I'alga in which the left spine is very reduced or al)sent may lie used in 

 conjunction with the name tropica in localities where the right spine is very elongate, and 

 where tiiese reductions occur, e.g., K. valga f. tropica-asyininctrica and K. 7'olga f. tropica- 

 motistrosa. 



The following terms may then be applied to designate the forms of A', zxilga. 



Keratella valga (Ehrenberg). 



Diagnosis: six anterior spines, reticulate sculpture forming a medial series of undixided 

 hexagons, maximum breadth of lorica very distinctly greater than the posterior breadth, the 



