228 Variations of the Lateral Shields in the Stickleback. 



niger is that the keel and a line continuing it along the back 

 to wlierc it joins the shield is the last part of the body to 

 retain the light ground-colour. Except in the totally black 

 variety (v. maura) L. cinereo-nifjer always has this keel and 

 line more or less evident as a thin pale stripe. 



This evidence will suffice to make it clear that, whatever 

 views we may hold as to the specific or subspecific rank of 

 L. cinereo-niger, it is not open for lis to acquiesce in the 

 statement that it " is simply one of the many colour- variations 

 of the well-known L. maximus, L." 



XXXVII. — Note on the Variations of the Literal Shields in 

 the Three-sjnned Stickleback (Gastrosteus aculeatus). By 

 G. A. BOULENGEK. 



Ever since Cuvier proceeded to divide the Sticklebacks into 

 species according to the presence or absence and the develop- 

 ment of the lateral armour, the question of the value of this 

 character has been much discussed. Most modern European 

 writers, with the exception of Blanchard and Sauvage, have 

 refused to accept Cuvier's species as such, although they 

 have usually retained them as varieties or subspecies. 

 These supposed species are, however, maintained provision- 

 ally by American authors, Jordan not long ago remarking 

 that he has not yet met with distinctly intermediate forms 

 either on the Atlantic or Pacific coast. Bonizzi, Day, and 

 Fatio have published results of investigations into the varia- 

 tions of the spines and shields in the smooth-tailed form from 

 one locality ; but the differences in the lateral armour in 

 northern brackish- water specimens do not appear to have yet 

 been subjected to a thorough statistical examination. 



In July last 1 collected indiscriminately in a tidal pool close 

 to Ostend Harbour sixty-six specimens of the three-spined 

 stickleback, with the object of testing their characters, as I 

 had observed that the three principal forms, viz. Q. trachurus, 

 G. semiarmatus, and G. gymnuriis, occurred promiscuously 

 both in and outside the harbour. The result is inter- 

 esting, as showing how complete the gradation between the 

 shielded and the smooth form is and how much the characters 

 may differ on the two sides of one and the same specimen. I 

 have therefore recorded the number of lateral shields in all 

 the specimens, and arranged them in a series from the most 

 perfectly armoured to the naked specimens. The numbers 

 given refer to the shields on either side, those of the left side 



