498 Mr. Jl, Gurnej on some British and 



related to A. cancriformis, while one of them (A. mauri- 

 tanicus) is founded on specimens already named by Daday 

 as A. cancriforniis. In separating these species, Ghigi has 

 relied largely on three characters : — 



(1) The presence or absence of teeth on the median 

 carina of the dorsal shield. 



(2) The form of the spines encircling the abdominal 

 segments. 



(3) Tlie armature of the last abdominal segment or 

 telson. 



The first two of these characters have not previously been 

 regarded as of great importance in discriminating species, 

 and it seemed to me worth while to examine such specimens 

 of ^. canc7nformis as I possess to see how far these characters 

 can be regarded as constant. My material, all of which I 

 have regarded as belonging to the true A, cancriformiSj is 

 as follows ; — 



(1) A number of mature, but small, specimens taken by 

 Mr. Balfour Browne from a pool on Preston Merse, Kirk- 

 cudbrightshire, in Sept. 1907 (see 'Nature,' Ixxvi. 1907, 

 p. 589). 



(2) Ten specimens, male and female, from near Tilghemt, 

 92 miles south of Laghouat in Algeria. 



(3) Two specimens from Laghouat. These and the pre- 

 ceding were taken by Dr. E. Hartert in April 1911. 



(4) A few females taken by me in March 1913 from a 

 pool by the Tunis-Bizerta road in Tunisia. 



With regard to the three characters named, the result of 

 comparison of this material is as follows : — 



(1) In all the Algerian and Tunisian specimens the dorsal 

 carina is perfectly smooth, but, whereas in the Tunisian 

 examples it ends posteriorly in a conspicuous spine, in those 

 from Algeria this spine is either quite small or entirely 

 absent. In the Scottish individuals the conditions are not 

 the same. Here the carina ends in a large spine projecting 

 into the posterior sulcus, and in nearly every case there are 

 one or more small incisions or teeth in its posterior part 

 (fig. 1,B, C). In this respect the Scottish examples diff'er 

 from the type and approach Ghigi's species A, maur it aniens. 

 In one or two cases (Tunis) I have found small denticles 

 scattered on the shield posteriorly (fig. 1, A). 



(2) I cannot follow Ghigi in attaching importance to the 

 differences described by him in the size and shape of 

 the marginal spines of the abdominal segments. Whether 

 tliese spines project or not beyond the edge of the segment 



