THE COLOEATION OF REEF FISHES 549 



one's shoulders, and air pumped into it through a hose reaching 

 a boat above continually replaces that which is exhausted and 

 escapes in turn below its rim. A total weight of about 80 

 pounds is necessary to steady the diver in 15 to 20 feet of water, 

 where excursions may be made up to at least half an hour, and 

 probably much longer, without discomfort. For the prosecution 

 of a research such as the present the hood has all the advantages 

 and none of the objectionable features of a regulation diving 

 suit. With a wax-coated slate and sharp-pointed nail ample 

 records may be kept and transcribed at leisure. 



Some of the fishes one encounters when diving are so little 

 afraid that they will almost feed from one's hand. Hence, even 

 minor changes in their coloration may be observed to advantage. 

 This encouraged an attempt at submarine photogi'aphy which 

 was at least as successful as might reasonably be anticipated 

 when working under novel conditions with untested apparatus. 



A No. Graphic camera, taking a picture 2f by 1^ inches, 

 was enclosed in a brass box, mndows in which permitted the 

 use of the finder, while the necessary manipulation of shutter 

 and film was accomphshed by means of a screw and plunger 

 protected by water-tight packing. Since the lens was of uni- 

 versal focus adjustment for distance was unnecessary. Some 

 of the pictures secured are reproduced herewith, and serve to 

 illustrate the striking changes in pattern which the fishes un- 

 dergo, and suggest also how slight ground there is for consid- 

 ering the animals conspicuous. Much remains to be done, but 

 in the light of present experience it is apparent that adequate 

 exposition of the facts should be possible by the photographic 

 method. 



Regarding color change it seems perfectly safe to go much 

 farther than Townsend ('10), who notes that his fishes showed 

 more color phases as their surroundings were more varied, i.e., 

 that their environment played at least a subordinate part in 

 determining their aspect; farther even than Mast ('14), who 

 states his con\dction that adaptive color changes in fishes occur 

 not only in marked degree, but that they are rather widespread. 

 There is, indeed, within the limits of my own experience almost 



