98 Journal of the Mitchell Society [Augnst 



one or two eggs, six is a high number, and nine was the maximum ob- 

 served in 1922. The mean number of eggs per brood pouch varied in 

 the several collections made from May 1 to December 12 from 1.5 to 

 4, but the means of actual numbers in the live Limnoria would have 

 been a little higher, for some eggs and larvae are lost from the brood 

 pouches in processes of preservation and shipping. In the early spring 

 collections of 1923, eight, nine and ten were not uncommon numbers 

 of eggs per brood pouch, twelve was the maximum observed (April 

 30) and the mean number per brood pouch for the collections of April 

 14, 21, and 30, and May 7, 15 and 23, Avere, respectively, 5.5, 5.5, 6.6, 

 4.2, 5.3, and 5. 



The salinity of the water alongshore of Fiver's Island varies to 

 some extent with the tide. There may be a difference in specific 

 graA'ity of .006 between high and low tides, although ordinarily the 

 difference is not more than .002 — .004. The variations from day to 

 day or season to season, according to conditions of rainfall, are much 

 greater. Thus the extremes found in Mr. Hatsel's records for the 

 period from May 1, 1922, to April 30, 1923, are 1.0232 for certain 

 days in January and February and 1.008 in August. Densities above 

 1.020 or below 1.012 may be sustained for several days, but the read- 

 ing of 1.008 occurred only once. The available observations indicate 

 only that exposure for several days to water of a density either as low 

 as 1.011 or as high as 1.0021 appears to have no unfavorable effect 

 upon Limnoria at Beaufort. 



Chapel Hill, N. C. 



