COTTIBvE. 47 



nostril tubular: the posterior patent: no barbels. Teeth fine in the jaws and in 

 a horse-shoe shaped band on the vomer : none on the palatine bones or on the 

 tongue. Fins first dorsal spines weak and about half as high as the rays of the 

 second dorsal: the two fins are usually slightly connected at their bases. 

 Ventral rays unbranched, the fin does not normally in British examples reach the 

 vent. Pectoral rays unbranched. Vent situated about midway between the 

 snout and the base of the caudal fin. Intestines the length of the intestinal 

 tract almost equals that of the entire fish excluding the caudal fin. Four ccecal 

 appendages. Scales absent : but in some examples the skin of the upper half of 

 the body studded with little rough points. Lateral-line about the middle of the 

 depth of the body or slightly nearer the back : it consists of about 33 tubes. 

 Ovi-sac black. Colours greenish-yellow, lightest beneath. Banded, striped, or 

 irregularly spotted with black: a light band, often of a V-shape across the 

 occiput. Fins orange : a dark band along the middle of the first dorsal fin : the 

 others with dark bands of spots. 



Varieties. The different varieties generally admitted to belong to this species 

 are divisible into : 



A. Pectoral fin, with 7 or 8 branched rays. France. 



B. Pectoral fin, without branched rays. 



a. Ventral rays branched. Scandinavia. 



b. Ventral rays simple. England, Germany, Austria. 



Cottns fozcilo'pus is stated to have D. 8-9 | 16-18, A. 13-14, the lateral-line 

 running above the middle of the height of the body and ceasing prior to reaching 

 the base of the caudal : ventral fin inserted midway between the snout and the 

 base of the caudal fin, while it reaches the vent, I have a beautiful example, 

 4^ inches in length, sent me by Professor Giglioli of the Florence Museum ; it has 

 D. 8/17, V. 1/4, all simple: 5 upper pectoral rays branched. A. 12. Vent midway 

 between the snout and the base of the caudal fin : lateral-line median and just 

 reaching the base of the caudal fin. Ventral fin nearly reaches the vent, at any 

 rate as closely as is seen in the examples of G. poecilojjus in the British Museum. 

 In Cuv. and Val. it is stated that the French G. gobio has from 6 to 8 or even 9 

 dorsal spines; the second dorsal 17 to 18 rays; and the anal 13. Blanchard, 1. c. 

 says of the French form it sometimes has 7 or 8 dorsal spines, that the second 

 dorsal fin has usually 17 rays : that in some examples all the pectoral rays are 

 simple, in others the upper 7 or 8 are branched ; while the anal has up to 13 or 14 

 rays. Unless my fish from the Florence Museum is considered a new species, it 

 must be admitted to be a link between 0. pozcilopus and G. gobio. Irrespective of 

 this, Girard in his elaborate monograph on the North American Cottoids, remarks 

 having seen "a mutilated specimen of an English miller's-thumb, with evidently 

 branched pectoral rays," thus showing, not that two species exist in Britain, but 

 that whether the pectoral rays are branched or unbranched, may occur in fishes 

 from the same locality, and must be regarded as a variable and inconstant sign of 

 no specific value. 



Respecting the situation of the vent, Richardson observed that in C. cognatus 

 " the anus situated precisely as in G. gobio, is rather nearer the snout than the 

 insertion of the caudal." One author states the vent in G. gobio to be rather 

 nearer the snout than to the tip of the caudal. Examining eight Gloucestershire 

 examples from 2 - 50 to 2 - 90 inches in length, captured the same morning, I found 

 their average entire lengths were 2'69 inches : and the average distance from the 

 snout to the vent 1*19, while the average length of the caudal fin was 043. The 

 vent in some was equidistant between the snout and the base of the caudal fin : 

 in others, between the snout and the first fourth of the caudal fin. Much the 

 same proportions are likewise present in the British Museum examples, whether 

 British, Swedish, Austrian, or from Dalmatia. In the example of G. gracilis in 

 the National Collection, 2 - 30 inches in length, the vent is 1T0 inches from the 

 snout, being very little different from what is seen in some similarly-sized 

 Gloucestershire specimens of G. gobio. 



As to the position of the lateral-line. In Leach's example, 2 - 90 inches long, 

 from Devonshire, the distance from the vent to the lateral-line is 0'45 inches, 



