448 



SCIENCE 



[N. S. Vol. XXXIV. No. 875 



Michael Sars North Atlantic expedition in 

 1910. Dr. Bigelow cites, as one of tlie impor- 

 tant results of the expedition, the experiments 

 undertaken to discover the depth to which sun- 

 light penetrates below the surface of the ocean 

 and the relation between this and the bath- 

 ymetrical distribution of animals exhibiting 

 certain types of coloration. Dr. Hjort found 

 that the red rays are absent and the blue and 

 violet raj's present at 500 meters; at 1,000 

 meters ultra-violet rays are perceptible and 

 at 1,700 meters no trace of light could be 

 detected. The black fishes and red prawns 

 taken in the daytime in temperate latitudes 

 were from a depth of 500 meters or more, i. e., 

 below the penetration of the red light rays. 

 In more northern latitudes these animals were 

 taken nearer the surface. In these regions 

 the red light rays probably do not penetrate 

 so far below the surface. Above the 500- 

 meter level the fishes were found to be " char- 

 acterized by lateral compression, larger and 

 often telescopic eyes, light organs and silvery 

 sides." These facts led Dr. Hjort to suggest 

 that the lower margin of the area penetrated 

 by red light rays marks the border between 

 two diiierently colored faunas. Dr. Bigelow, 

 in support of this view, states that the Me- 

 dusae apparently can be divided into two color 

 groups which overlap at 250 to 300 fathoms. 

 The species above this level are characterized 

 by little pigment and iridescence, and those 

 below by red and brown pigment. 



The same relation between the vertical dis- 

 tribution and the coloration was found to exist 

 in the young. The young of some of the spe- 

 cies spend their larval existence near the sur- 

 face and do not exhibit the adult coloration, 

 this being acquired as they increase in size 

 and descend to the habitat of the adult. The 

 young of other species develop in the same 

 region in which the adults are found and 

 acquire the adult coloration much earlier. 

 The coloration and vertical occurrence are cor- 

 related from the earliest stages. 



The writer, during the preparation of a 

 monograph of the Cyclogasteridffi, has had 

 occasion to trace out the correlation between 



the coloration and the bathymetric distribu- 

 tion of these fishes. The results obtained are 

 of interest in connection with those obtained 

 by Dr. Hjort and Dr. Bigelow and it seems 

 opportune to present a general account of 

 them in advance of the main body of the work. 

 In reviewing the results presented here it 

 should be borne in mind that the records upon 

 which they are based are very incomplete 

 when compared with those available to Dr. 

 Hjort. It should also be noted that Dr. 

 Hjort's conclusions result 'from the study of 

 the general fauna while those of the writer 

 are based upon the examination of a single 

 family. This may account for the difference 

 in the conclusions arrived at. The methods 

 employed by the Albatross are very unsatis- 

 factory for the solution of problems dealing 

 with vertical distribution. The dredge is sent 

 down and haviled up open, catching forms 

 through all the intervening depths. Unless 

 the animals captured have some peculiarity of 

 structure which indicates their habitat as be- 

 ing on the bottom it is impossible to decide at 

 what depth they entered the dredge. The in- 

 termediate hauls are made at a depth of 300 

 fathoms and the net hauled up open. The 

 absence of species from the intermediate hauls 

 indicates that their habitat is below this 

 depth, but how far below remains a mystery. 

 Also records of the coloration of the fishes, as 

 they first appear, are very seldom taken. The 

 colors frequently change in spirits. The 

 translucent reddish cyclogasterids usually be- 

 come an opaque white. This restricts the 

 conclusion that can be drawn from the study 

 of such specimens. 



The Cyclogasteridse is a favorable group in 

 which to work out the modifications of struc- 

 ture and color as the species become adapted - 

 to the deep sea. This is true because the 

 family is abundantly represented at all depths 

 from the tide pools down to 1,973 fathoms. 

 About 42 species are known to inhabit depths 

 of less than 100 fathoms, 49 inhabit the region 

 between 100 and 500 fathoms and 34 the 

 depths greater than 500 fathoms. Starting 

 with the tide-pool species as the most primi- 

 tive, we can readily trace out definite niodifi- 



