646 



SCANDINAVIAN I'ISHKS. 



ou tlie other hand, being continent witli the 2n(i (fig;. 

 l(i:i). 1 specimen in (il. 



I IV II (J III 



1—2—3—4—5—6—0—8—9 

 i: uitli n supcrnumerarv spinous vay on the 2nd inter- 

 spinal plate, otherwise similar to <;. 1 specimen in fil. 



\\ I _I1_ III 



1 — 2 — 3 — 4 — 5 — 6 — (I -.s — 9 



C: in one Three-spined Stickleback with a rudiment of a fifth 

 spinous rav and with eight interspinal plates. 

 O" 1__ I^ \ IV III 



1—2—3—4—5—6—7—8—9 



Now, as at least 7 of the interspinal plates — 

 though only 3 — 5 at once — may be furnished in the 

 Three-spined Stickleback with true spinous rays, and 

 as 7 — 12 free s])inous rays occur at the dorsal margin 



t/s.3 



ds.^. 



Fig. 162. Gastcrosteus actileatns, Iraclntms, tetracanthus (sjiinnlosiiis) 

 with tlie siiijerniimerary spinous ray (') set on llie posterior, de- 

 tached part of the third dorsal phite. Natural size. From the Frisuhes 

 Haif, off Konigsberg. After Heinoke. 



of the Ten-spined Stickleback, while in the Fifteeu- 

 spined Stickleback the limits of the variation in this 

 respect are 13 and 16, the series, with decrea.siiig num- 

 bers, is complete from the last species to that first 

 mentioned. Heincke assumes that the development has 

 taken this direction, with constantly sinking numbers. 

 Thus, where more than 3 free spinous rays occur at 

 the dorsal edge of the Three-spined Stickleback, we 

 have a retrogression to an older stage of development, 

 through which the species has already- passed. The 

 Fifteeii-spined Stickleback, according to this theory, re- 

 presents the earliest stage of the development of the 

 whole family, whose descent should thus be traced 

 from fishes of the Fistularioid family. 



The alterations of development, however, lead us 

 to a different conclusion, namely that the Ten-spined 

 Stickleback stands nearest the probable origin of the 



famil\ . \\ c have already .seen thnt the elongated furni 

 Avhich is the chief characteristic of the genus (jafitrri'd 



— though we meet with it in the youngest forms of 

 all the Sticklebacks — especially in the character th;it 

 suggests a transition to the Flute-mouths, nanieh- the 

 elongation of tlie snout, is a character of age, a cha- 

 racter which develops during growth from a iu\enile 

 stage with the typical form of the species perfect in 

 other respects, but with the length of the snout only 

 slightly greater than in the true Sticklebacks. Even 

 during the later (older) stages of growth this general 

 rule holds good, for the length of the snout shows 

 increase, c. g. in proportion to the po.storbital length 

 of the head (see the following taljle), both in the Fifteen- 

 spined Stickleback and the true Sticklebacks, but is least 

 in the Ten-spined species. Another relation, in which 

 the changes of development also follow the same direc- 

 tion ill all the three species now tuider consideration, 

 gives the same result. If we endeavour to formulate 

 an expression for the position of the soft-rayed dor.sal 

 and the anal fins with respect to each other, we find 

 that, as a general rule, the anal sjiine lies somewhat behind 

 the perpendicular from tlie beginning of the former tin 



— in the Ten-spined Stickleback alone it may e.xcep- 

 tionally lie somewhat in front of this line — and that 

 the percentages for the distance between the former fin 

 and the tip of the snout compai'ed witli that lietween the 

 anal sjiine and the same point, run in all three species 

 in inverse proportion from the earlier stages of growth 

 to the later. The highest percentages, \vhich are thus 

 expressions of the lowest stages of develo])ment, occur 

 in the Ten-spined Stickleback (see the following table). 

 But this is not all. In the relations in which the Fifteen- 

 spined Stickleback shows a deviating direction of deve- 

 lo])ment, it stands nearest to the Ten-spined Stickleback. 

 For example, if we compare the length of the soft-rayed 

 dorsal fin with that of the head, we find that this fin 

 is considerably longer in the true Stickleliacks than in 

 Gasfro'd, and longest in the Three-spined Stickleback. 

 The percentage for this relation also increases, as a ge- 

 neral rule, during the growth of the Sticklebacks, but 

 decreases with age in Gasfrrea (see the following table). 



Length of the hodv expressed in millimL-tres G4.9 



„ ,, Huout in % of the postorbitftl length of the head %.l 



Distance between the soft-riiyed dorsal fin and the tip of tlie snout in ?o of the distance between the anal | 



Bpine and thf same point I 97.2 



Length of the base of the soft-rayed dorsnl tin in % of the length of the head | 52.6 



