250 CALIFORNIA ACADEMY OF SCIENCES. 



The specimens here noted agree with the original de- 

 scription of that species, except in the presence of plates 

 as above noted. The original description was based on 

 eight specimens. Our specimens from the San Miguel 

 Creek have an average of four plates, only one having 

 none. Adobe Creek furnishes 115 specimens, with from 

 four to six plates. None of the 175 specimens from San 

 Francisquito Creek have fewer than four plates, and four 

 specimens are fully plated. The specimens from Walla 

 Walla River range from four plates to a fully plated body, 

 the average number of plates being fifteen. Seventy-nine 

 per cent, of the specimens from Pilarcitos Creek, empty- 

 ing into Half Moon Bay, are fully plated, but some have 

 as few as six or seven. From several localities the spec- 

 imens are all fully plated, but it is worthy of note that in 

 such cases we have only a few specimens. Larger col- 

 lections would doubtless show variation. 



Dr. G. A. Boulenger has recorded similar variations in 

 the English Stickleback, Gasterosteus aculeatus. The 

 variations described by him were all found in one local- 

 ity, in practically one haul of the seine. They corre- 

 spond to those of our specimens from the Walla Walla 

 River (see table). 



From a number measurements, made with as much 

 accuracy as the small size of some of the parts measured 

 would permit, I find that the dorsal and ventral spines 

 become shorter, the prepectoral area becomes slightly 

 smaller, and the pubic plate shorter as the number of lat- 

 eral plates grows less. There is little variation in the 

 size of the head, the width of the thoracic area or the 

 depth of the body. The proportionate size of the eye 

 varies inversely with the size of the specimen, as usual 

 in fishes. These variations are all such that those found 

 in one locality overlap those found in another, thus form- 



