(e.g., Lyopsetta exilis) which we did not reexamine 

 systematically. 



COMPUTER ENTRY AND EDITING 



Each taxon on the original identification sheets was given a 

 3-digit code based on the list of codes in Haight et al. (1979). 

 Taxon codes and counts from these sheets were keypunched by 

 cruise and station, along with pertinent station and tow data and 

 entered into the VAX 11/780 computer at the University of 

 California, San Diego Computing Center. After entries were 

 completed for an entire year, print-out listings of taxa and 

 counts on each station were compared with the original data 

 sheets to eliminate keypunch errors. Next, data in the file were 

 cross-checked with data on an existing file which contained: 

 station and tow data; numbers of eggs of sardine, anchovy, and 

 saury (Cololabis saira) ; numbers of larvae of sardine, anchovy, 

 hake, jack mackerel, and Pacific mackerel; total number of fish 

 eggs; and total number of fish larvae. 



Discrepancies in ichthyoplankton data in these two files 

 were corrected by inspecting original records from the sorting 

 laboratory, the original ichthyoplankton identification sheets, 

 and the samples themselves. Station and tow data discrepancies 

 between the two files were corrected by reviewing ships logs and 

 deck tow sheets, original records from the sorting laboratory, 

 cruise announcements, publications, header information on the 

 ichthyoplankton identification sheets, and station plots 

 generated for each cruise. Eventually all station and tow data 

 were checked by comparing these sources. 



The corrected ichthyoplankton data base was then examined 

 statistically and outliers were found and checked as above. 

 Distributional plots were then prepared for each taxon and these 

 were checked by reviewing the data sources mentioned above and by 

 examining archived specimens. A listing of each taxon by station 

 (Table 4) was produced, which became the primary document for 

 subseguent checks. Misidentif ications found in geographic 

 outlier checks and other misidentif ications and data problems 

 discovered in the course of examining archived samples resulted 

 in several iterations of Table 4. Finally, totals in Table 4 

 were checked against annual summaries of incidence and abundance 

 (Tables 2 and 3). Ecological analyses of the data (Moser et al., 

 1987) were conducted concurrently with editing procedures and 

 provided cross-checks that allowed correction of errors. 



SPECIES SUMMARY 



Larvae of Pacific hake (Merluccius productus) represented 

 24% of all fish larvae taken on CalCOFI cruises during 1951 and 

 numbered twice as many as northern anchovy (Engraulis mordax) , 

 the next most abundant species (Tables 2, 3). Hake incidence was 

 comparatively low (11th) as was anchovy incidence (7th) , 



10 



