2. SYNGNATHUS. 159 



/S'. CHvieri is described as having the " snout half the length of the 

 head. Eleven or twelve rings before the dorsal fin, of which two 

 belong to the body. D. 38-41." 



However, these apparently important differences are nothing but 

 iiiarcuracies of the describer, who gives the measurements of indi- 

 viduals contradicting the statement of the diagnosis, and showing 

 that the relative length of the snout is precisely the same in both. 

 Why he states that " two rings belong to the body," whilst, in 

 fact, all the eleven or twelve rings before the dorsal fin are body- 

 rings, is not clear. There is an error or omission also in this sen- 

 tence. The number of twenty-two dorsal rays in S. cir/assizii = buc- 

 culenhis (llathke, spelt luculentus bj' Dr. Kaup) is not in accordance 

 with those given by previous writers, Eathke stating 35-30, and 

 Nordmann 35. 



Having such short-snouted examples from the North Sea, Medi- 

 teranean, and Black Sea, I have come to the conclusion that they are 

 young S. acus. 



However, it is very singular that young examples (from 4 to 

 7 inches) differ so much from old ones (8 to 18 inches), that, if the 

 differences were observed in examples of the same size, they would 

 be considered to be specific. Beside the shortness of the snout, the 

 number of osseous rings of the body is less (15-19), frequently 

 also that of the dorsal rays (sometimes as low as 31), and the 

 lateral line is more often continuous with the upper caudal edge 

 than interrupted. 



In the following list I intend to show the great variability of these 

 examples : — 



inch. D. Oss. rings. Lat line 



11. Firth of Forth 3? 4^-6^ 36-40 lG-t-42or43 contin.' 



\2. „ „ 1 (? 5 40 17-1-40 contin. 



3. England 1 9 4^ 43 1C-|-41 interr. 



4. Wales 1 c? 5 35 10441 interr. 



5. Bohusliin IJ 6 37 17-1-42 interr. 



0. „ 1 S 5 39 17-1-4(1 contin. 



7. „ 1 H 5 42 17+41 contin. 



8. Gibraltar 2? 4 31-34 1S+.38 contin. 



{». Black Sea 5 S& $....5-GJ 31-34 16+38 contin. 



10. Trieste 1 S 4* 31 17+35 contin. 



(Type of S. hrevi- 



rostris.) 



11. Sicily 1 ? Oi 37 19+42 contin. 



12. ? 1 ? 6 40 16+42 contin. 



13. ? 1 <? 4 40 15+38 interr. 



Although I have examined a good many more young examples 

 than those mentioned, I have never met with a single Si/nr/iKtiJuix 

 under 8 inches in length which agreed entirely with a S. acus above 

 that size. And stiU more curious is the fact that these fishes (which 

 cannot be considered to be fully developed) are capable of propnga- 

 ting their species, the ovaries of the females and the pouches of the 

 males being filled with mature ova. 



