94 Papers from the Marine Biological Laboratory at Tortugas. 



While the amount regenerated in sea-water was equaled or exceeded in 

 solutions diluted to as much as 85 per cent, the optimum solution was 95 

 per cent sea-water. With increasing concentration, regeneration fell off 

 rapidly; with increasing dilution the amount regenerated was diminished 

 very slowly. 



During the succeeding 10 and 16 days there was a relatively greater 

 regeneration in solutions below the optimum, i. e., in the 75 and 70 per cent, 

 and the character of the curve was correspondingly altered. 



The development of supernumerary arms likewise occurred most 

 frequently in the optimum solutions, namely, 95 to 85 per cent; less fre- 

 quently with increasing dilution or concentration. 



When these results are compared with those of Loeb on Tubularia and 

 with the writer's on Eudendrium, it is clear that in all the maximum regener- 

 ation does not take place in sea-water, but in sea-water diluted, and that 

 with increasing concentration or dilution regeneration is retarded and 

 diminished. 



The character of the curves representing the influence of changes in 

 salinity upon the regeneration of these organisms is strikingly different. 



1. The most dilute solutions in which regeneration will take place is 

 40 per cent for Tubularia, 50 per cent for Cassiopea, and 60 per cent for 

 Eudendrium. 



2. The maximum regeneration occurs In 95 per cent solutions for 

 Cassiopea, while it is 85 per cent for Eudendrium and 65 per cent for 

 Tubularia. 



3. The regeneration in sea-water is equaled or exceeded in solutions 

 diluted as much as 13 per cent for Cassiopea, 20 per cent for Eudendrium, 

 and 50 per cent for Tubularia. 



4. The curve is most asymmetrical for Tubularia, less so for Cassiopea, 

 and almost symmetrical for Eudendrium. 



5. The Eudendrium behaves more like Cassiopea; both are strikingly 

 different from Tubularia. 



It is very evident that Loeb's curve is not representative of the behavior 

 of marine organisms to changes in concentration, and it is altogether prob- 

 able that such a curve can not be expressed in a simple curve based on 

 two variables. 



In the absence of information concerning the exact concentration of the 

 sea-water, where each of these experiments was made, it is unprofitable to 

 make extended generalizations. 



