1923 ] Esterly: Marmt Copepoda at La J oil a 421 



genera is now in contemplation. In addition to the forms mentioned, 

 several other species occurred rarely and in small numbers, so that 

 their presence need only be mentioned. 



When the project of taking such an extended series of observations 

 was being talked over, we felt that a tabulation of results would help 

 in answering such questions as the following: (1) Is there a seasonal 

 variation in abundance of copepods ? If so, are the times of maximum 

 abundance similar or different for different forms. 1 '2) Are there 

 diurnal (night and day) variations in abundance? If so, are there 

 generic or specific differences as to the times of greatest abundance ? 

 (3) Is there any relation between variations in numbers of organisms 

 and changes in such conditions in the water as temperature, for ex- 

 ample, or salinity or acidity ? 



The results we have arrived at thus far partially answer the first 

 two questions, but correlations between abundance and the other 

 factors mentioned have not been worked out as yet. 



In planning the work the intention was that E. L. Michael should 

 handle the statistical part of the results. Before his death in 1920 

 he had gone over some of the collection reports and had dealt with 

 them by methods developed by him in collaboration with McEwen. 

 Michael's findings were never written up, however, and since his mode 

 of working requires special knowledge of statistical procedure, it has 

 seemed desirable to report on the collections without employing greatly 

 refined methods of handling the data. Those of the present staff 

 who have consulted on the matter believe that even these cruder 

 tabulations show in a general way what has been gained during the 

 two years, and, equally important, show also what has not been gained 

 and whether the plan should be continued with or without modifi- 

 cations. 



The method of computing used in preparing this summary was 

 suggested to me by McEwen. The computations start with the average 

 number of animals per haul, arranged under headings for each of the 

 twenty-four months over which the observations extended and for each 

 of the six times of day at which the collecting was done. Such an 

 arrangement of the original data for one species is shown in table 1. 

 The average per haul for the two years (last line of table) was used 

 as a standard, and two methods were followed which may be called 

 respectively the '"ratio of the means" and the "means of the ratio." 

 By the use of these methods coefficients are obtained which make it 

 possible to correct the original entries, so that in dealing with the 



