422 



University of California Publications in Zoology [Vol. 



time of day relations the seasonal variability is eliminated and an 

 estimate may be made of the reliability of the results which relate to 

 the time of day. 



TABLE 1 



Average Number op Acartia tonsa per Haul, Arranged According to 

 Months and Times or Day 



Year and month Sam. Noon 4 pm 8 p.m. Midnight 4 a.m. 



1916 



IX 26 12 12 49 198 82 



X 4 3 5 37 41 23 



XI 40 12 7 74 53 130 



XII 24 14 19 19 20 32 



1917 



1 9 4 19 5 5 10 



II 11 1 1 5 5 11 



III 16 5 47 80 83 207 



IV 37 10 9 46 59 63 



V 4 1 34 93 63 121 



VI 5 1 1 82 148 38 



VII 11 16 97 103 65 



VIII 29 36 123 237 304 168 



IX 25 2 334 270 60 220 



X 14 2 1 43 338 298 



XI 35 4 9 32 107 105 



XII 11 13 42 85 66 109 



1918 



1 8 10 3 58 75 69 



II 65 8 15 28 



III 108 1 43 28 29 



IV 0.3 4 18 58 26 39 



V.... 20 10 5 95 76 21 



VI 3 12 101 67 73 



VII 2S8 947 469 915 1026 1406 



VIII 41 24 26 79 175 258 



Average 35 47 50 119 131 150 



To give an actual example of the way in which the data were 

 handled we may use the figures in table 1. The method of "ratio of 

 the means" is as follows: (1) Add the average numbers of animals 

 per haul at the six periods for the two years. (2) Divide the sum 

 obtained by the horizontal sum of the entries for each month ; the 

 quotients are then the coefficients. (3) Multiply each of the original 

 entries (table 1) by the coefficient for the corresponding month; the 

 products are the values corrected for season. 



The method called the "means of the ratios" is as follows, refer- 

 ring again to table 1. (1) The ratio of each entry in a column to the 



