54 THE AMERICAN ARBACIA 



each other. It is quite essential that they are not overcrowded (150- 

 200 in an average sized aquarium, 16 gallons), that there is a steady 

 flow of salt water, and that the water is kept unpolluted. Any unfavor- 

 able condition will cause the animals to shed, and when one sheds the 

 others are likely to shed also. Two lots of six animals each were kept, 

 one set in the light and one set in the dark from July i to Sept. 16, 

 1948, and all were in perfect condition at the end of that time. 



The eggs from animals kept for some time in the aquaria are slightly 

 different late in the season, taking longer to cleave, irrespective of tem- 

 perature, and they are more viscous, taking longer to stratify and break 

 apart with centrifugal force ; also some of the red pigment granules 

 remain in the white half. The rate of oxygen consumption is decreased 

 (Shapiro, 1935 c). The amount of carbohydrate and of phosphorus is 

 greater and of nitrogen is less in the stored eggs (Hutchens, Keltch, 

 Krahl, and Clowes, 1942; Crane, 1947). It seems probable that there 

 are other differences of a physiological or chemical nature, but for 

 many purposes the eggs are quite suitable to use as they develop nor- 

 mally and give normal plutei. A difference between summer and winter 

 eggs has been observed for Paracentrotus lividus at Naples (Horstadius, 

 1925). Some eggs obtained from A. punctulata which had been kept 

 at Woods Hole until the middle of January appeared normal except 

 that they were pale. These, when fertilized, developed to blastulae 

 (A, E. Navez, personal communication, 1936). Good sperm has been 

 obtained from animals kept over winter until March; the females 

 from this lot had full ovaries, but only degenerate and abnormal eggs 

 which had apparently not been shed the previous season (E. B. Harvey, 

 1942). 



It would seem that in general at Woods Hole under natural con- 

 ditions, A, punctulata shed their eggs, when fully ripe, the latter part 

 of the summer, then do not have ripe eggs again until the following 

 summer. It is quite possible, however, under laboratory conditions for 

 a ripe animal to shed some of its eggs at one time and more later on. 



The breeding season of A. punctulata at Beaufort, N. C, is con- 

 siderably in advance of that at Woods Hole. They are fully ripe and 

 in fine condition during May (E. B. H.). Some animals have ripe eggs 

 and sperm in January (C. B. Metz, personal communication, Jan. 

 1951). In Florida, at Alligator Harbor, they are ripe from September 

 until June and in fine condition in February, March, and April (C. B. 

 Metz, personal communication, Jan. 1955). 



The closely related species, A. lixula, which is common at Naples, is 

 ripe all the year according to Lo Bianco (1909), but the eggs are bad 



