1898.] beeedhstg of the db,a6oto;t. 287 



developed secondary characters. One of these was found to 

 measure 24'7o cm. (9| inches). The other appeared to be almost 

 exactly the same size. A female, one of the largest, was 17"15 cm. 

 (6| inches) long. There were also a number of fish of about 

 the same size and smaller, including both females and young 

 males. 



Previous to the 11th February the Dragonets were sluggish 

 in disposition, frequently burying themselves in thegi'avel, though 

 active enough at feeding-rime. The large males were not observed 

 to display their dorsal fins, though young fish will often raise the 

 first dorsal. The colours of the large males, so far as they could 

 be seen, were not remarkably brilliant. 



Entering the aquarium at about 9 a.m. on the morning of the 

 11th February, I noticed a pair of Dragonets ascending together 

 to the surface of the tank, and found, on observation, that 

 pairing was in full progress. The operation was repeated every 

 day until the 19th February, so that 1 had ample opportunity 

 of noting the details, which- 1 shall attempt to describe. 



While pairing is in progress all the Dragonets in the tank 

 appear to be in a state of great excitement, especially the two 

 large males. These keep darting along the bottom of the tank at 

 short intervals, at the same time exhibiting all their finery. 

 Sometimes the dorsal fins are erected before the fish starts, 

 oftener at the instant of starting, while the mouth is protruded 

 to its utmost, causing the roof of the gi'oove which lodges the 

 ethmoid process of the premaxillse to be raised nearly or quite 

 to the level of the top of the eyes. The giU-covers are inflated 

 and the hyoid apparatus is depressed, while the pelvic fins are 

 held rigidly forward and outward. The attitude is well shown 

 in my sister's drawing (Plate XXVI.)\ The fish scarcely leaves 

 the bottom, the anal tins remaining depressed and out of sight. 

 The motive power is furnished by the pectoral and caudal fins, 

 sometimes by a stroke of the whole tail. The yellow bands are 

 much more brilliant than they were noticed to be before sexual 

 activity commenced. They undergo no change with the elevation 

 of the fins, but retain their extreme brilliance only for the first 

 few days. The blue bands of the side flash out with intense 

 brilliance as the fins are hoisted, but become paler again before 

 they are lowered. This vividness of colour is attained only during 

 the first few days of sexual activity. 



The bhie bands of the head are but Httle, and those of the 

 pelvic fin not at all affected when the dorsals are raised. The 

 bands of the latter never attain the same depth of colour as those 

 of the side. Their colour is rather that of a turquoise, while 



1 Van "Wright's figure (H. Smitt, Hist. Scand. Fish. ed. 2, pi. xiv.) was 

 painted from life, but the depressed condition of the eyes and the position of 

 the pelvic and anal fins suggest that the outline was taken from a dead 

 specimen artificially arranged. My sister's drawing is a compilation of 

 sketches from life, cheeked by measurements of a dead specimen of about the 

 same size as the living model. ' 



