Variability of Bacteria i?i North Atlantic Sediments 



531 



o 



50 



r r 0.62 



COLONIES PER GRAM DRY WEIGHT 

 FRESH SEDIMENT 



Fig. 4. A comparison of bacterial counts made immediately at sea with 

 counts made six months later in the laboratory upon duplicate sediment 

 samples at -20 C since collection. The points are means of four filter 

 counts upon each of the duplicates. The solid line is calculated by method 

 of least squares; it does not differ greatly from the broken line indicating 

 complete agreement between the two counts. The correlation co-efficient 

 (r) indicates a probability of 0.01 for the twenty-one samples. 



Comparison of Counts Upon 



Fresh and Frozen Marine Sediments 



A simple method of bringing sediments back to tlie laboratory 

 without changing their bacterial count would be most useful. 

 To check the possibility that freezing might serve this purpose, 

 duplicates of most of the samples from the Labrador-Scotian 

 shelf were frozen and held at -20° C for o\'er six months before 

 counting. Figure 4 shows that these duplicate filter counts parallel 

 one another surprisingly well. Freezing and thawing of terrestrial 



