January 22, 1904.] 



SCIENCE. 



139 



brightly colored, forms .ar,e>.tliei'eby ren- 

 dered correspondingly more conspicuous 

 and, therefore, more liable to attack from 

 enemies. May it come within the category 

 of 'warning' coloration, due to the offen- 

 sive cnidarian armor borne by most of the 

 members of this phylum? So not a few 

 who have essayed an account of the matter 

 would have us believe. It seems to me, 

 however, open to serious doubt, aside from 

 the fact that it lacks evidence. On the 

 other hand, among hydroids I have found 

 that those having brighter colors are most 

 liable to be eaten by fishes in the habit of 

 feeding upon such a diet. Furthermore, 

 various worms, snails, etc., which are 

 known to feed upon them would be more 

 likely to be attracted by colors than to be 

 repelled. It is also matter of common ob- 

 servation that such animals are much more 

 abundant ainong colonies of highly colored 

 hydroids like Eudendrium, Pennaria and 

 Tuhularia than among species of Ohelia or 

 others of little color distinction. Many 

 fishes with finely adapted dental apparatus 

 are constant feeders upon corals, tran- 

 quilly browsing among the animated 

 foliage of this luxuriant forest. 



Finally, may it come within the cate- 

 gory of ' sexual selection ' 1 So far as I am 

 aware, no one has ventured to assign to it 

 any such a significance. Where sex char- 

 acters are so little differentiated as among 

 at least a portion of the phyhim such an 

 explanation would be as far-fetched as it 

 would be unnecessary. While upon the 

 part of some of the older naturalists there 

 was a disposition to regard the massing of 

 members of the Scyphomedusffi at certain 

 times as having a sexual meaning, it may 

 be doubted whether it has any consider- 

 able support in facts. 



Concerning coloration among the antho- 

 zoa, Duerden, whose work on the group is 

 so extended and so favorably Imown, has 

 summarized the following account: 



" The prevalence of the yellow and 

 brown color is, easily understood when an 

 examination is made of the polypal tissues. 

 For in all instances in which it occurs, the 

 entoderm is found to be crowded with the 

 so-called 'yellow cells' or Zooxanthellse, 

 which are unicellular, symbiotic algse, the 

 chromatophores of which are yellow or 

 yellowish-green. That these are the main 

 cause of the external coloration may be 

 easily proved from colonies of Madrepora. 

 In this genus the polyps toward the apex 

 of branches are nearly colorless, and on a 

 microscopic examination of the entoder- 

 mal layer Zooxanthellte are found to be 

 absent while they are present in abundance 

 in older pigmented regions." 



These symbiotic algse are not, however, 

 the only source of color among the corals. 

 Duerden finds ectodermal pigment gran- 

 ules, aggregated in somewhat irregular, 

 isolated patches in some cases, in others 

 somewhat regularly distributed. 



He also found that a third source of col- 

 oration among corals was the presence of 

 what he has termed 'boring alg£e.' These 

 were both red and green, and penetrate 

 into the skeletal mass and color it a dis- 

 tinct red or green, as one or the other may 

 be present. 



In his work on the Actiniaria of Jamaica, 

 this author has found in many cases the 

 presence of unicellular green alga growing 

 upon the surface and giving to the poljrp 

 a distinctively green color. He found also 

 superficial granular pigments in certain 

 species which could be removed by any 

 erosion of the ectoderm. I have found the 

 same in several species of New England 

 actinians, and in some cases the pigmenta- 

 tion was irregularly distributed, sometimes 

 in blotches, sometimes in longitudinal 

 stripes, more often the latter. So ex- 

 tremely variable is the coloration in many 

 of these organisms that it is impossible to 

 utilize it as a factor in differentiating spe- 



