DEPARTMENT OF MARINE BIOLOGY. 209 



agreement in habit, while among the others (which include many accepted as 

 examples of warning or immunity color) red is almost unknown even as a minor 

 element in a color scheme. 



The facts regarding red are in harmony with a general rule which is gradually 

 becoming apparent, that coincidence in habit is correlated wdth agreement in 

 pigmentation. Of 18 species (5 red ones included) which do not range over the 

 open reef freely by day, not one has a conspicuous graj'- phase, nor is one 

 marked with the sandy gray color of barren bottoms. Conversely, of 29 species 

 which do range over the reef comparatively freely by day, 17 have a gray 

 phase or permanent gray markings constituting an important element of their 

 pattern, and 10 show definitely, though in varying degrees, a color adjustment 

 converging toward that of the gray bottom, when they swim over or come to 

 rest upon it. 



Additional evidence that there has been misconception regarding the func- 

 tion of color appears from a consideration of certain fishes which are marked 

 with blue-gray or blue. Abudefduf saxaiiUs is such a one, conunonly con- 

 sidered conspicuous by observers viewing it from above. However, when it 

 is seen from a lower level its contour is visible at times not more than 15 feet 

 from the observer, for its blue-gray markings are resolved into the blue-gi*ay 

 haze which at that depth surrounds one on all sides and in lateral view 

 nothing is seen but several parallel bars of brown and yellow. The chub 

 {Kyphosus sedatrix) and the runner {Caranx ruber) may be almost invisible to 

 an observer at the same or a lower level than they are themselves. The limita- 

 tions imposed by time prevented the pursuit of this portion of the investiga- 

 tion to its logical conclusion, but the suggestion that shades of blue may be 

 assigned a definite significance is one of the most interesting that appears at 

 present. 



With regard to aspects of the general problem as yet of subordinate interest, 

 it may be stated that facts bearing upon the phylogeny of the color patterns 

 of the fishes continue to accumulate and that the inadequacy of the color de- 

 scriptions of the tropical fishes becomes more apparent as familiarity with their 

 color changes increases. As an example of confusion resulting from incomplete 

 knowledge of color change, it is to be noted that Bodianus fulvus and B. punc- 

 tatus are two color phases of one species of which one may almost instantane- 

 ously replace the other. 



Preliminary Report on the Comparative Morphology of the Recent Madreporaria 

 around Tortugas, by George Matthai. 



During the short stay I made at Tortugas (July 17 to August 1) my work 

 was limited to general observations on the coral fauna and coral reefs of that 

 locality. With the facilities at the Laboratory, it was possible to make an 

 extensive collection of all the common species for subsequent examination of 

 their soft and hard parts. I hope to commence the investigation of this ma- 

 terial after my return to Cambridge, England, at the end of 1915, and finally, 

 to incorporate the results with my studies on the Comparative Morphology 

 of the Madreporaria, which were begun at Cambridge University in 1911, 

 mainly on collections made by Professor J. Stanley Gardiner, during his expedi- 

 tions to the Indo-Pacific ocean. Such a comparative study will, it is hoped, 

 give us a better understanding of the anatomy and relationships of the corals 

 of the Indo-Pacific and Atlantic regions. 



All the common forms of coral were kept under observation in the Labora- 

 tory. The only species that extruded larvae was Favia fragum; none of these, 

 however, settled during the course of the fortnight I remained at Tortugas. 



