480 AUGUSTA LAMONT ON 



wards, and backwards, and covering the region between the propterygium, the outer 

 anterior border of the pectoral fin, and the side of the abdominal cavity. The canals 

 are, however, not subdivided into separate groups, but form a continuous series, 

 diverging over the region described, and ending at varying distances from the ampullae 

 in such a way that the area they cover forms a definite and characteristic shape 

 which varies slightly according to species. 



(5) Mandibular Canals. 



The canals of the mandibular ampullae are uniformly arranged in one group 

 opening on the ventral surface, and they are directed inwards towards the median 

 line, just posterior and parallel to the lower jaw. Not only is the capsule, which in 

 all cases encloses these ampullae, smaller than the others, but the number of its 

 canals is less, and these and their ampullae are smaller in size than those of the 

 other ampullary groups (Plate III, fig. 4). 



4. Distinguishing Features of the System in the Several Species. 



Of the various British species of Rata collectively the most noticeable point is 

 that they at once fall into two groups: — {l) Those with many, widely distributed, 

 and pigmented canals. (2) Those with fewer canals of more restricted distribution, 

 and without pigment. 



The first group contains only three species— Raia batis, R. macrorhynchus, and 

 R. oxyrhynchus. Of these Raia batis only was included among the species examined 

 during the present investigation, but it is obvious from Muller's description of 

 R. oxyrhynchus (which he refers to under the name of R. vomer) that it belongs to 

 this group. Couch's reference to the pigmented tubes appearing as dusky lines and 

 spots on the ventral surface of R. macrorhynchus is also conclusive with regard to 

 that species. 



The second group contains the remaining seven species, viz. Raia alba, 

 R. fullonica, R. radiata, R. circularise R. clavata, R. microcellata, and R. maculata, 

 of which only the first was missing from among the subjects of this investigation. 

 Among these there is a remarkable similarity in the general distribution of the 

 canals, the most distinguishing feature between the species occurring in the 

 number, length, and distribution of the posterior dorsal group of canals of the 

 hyoid ampullae. 



Following out Garman's idea that the canal systems may be used as a basis for 

 classification, one is at once led to the conclusion that Day's division of the genus 

 into "long-snouted" and "short-snouted" species is artificial, since that grouping 

 entails the classing of Raia alba and R. fullonica along with the types having 

 pigmented canals. Couch considered the grey colour on the ventral surface (due to 

 the presence of pigmented canals) of systematic importance, and went so far as to 



