7 Moderately strong (18-20 knot) north winds in the north section gave way to light (5-10 knot) northwest winds in the southern sections. Only 28 July was favorable for skiff work. Synoptic Highlights 1.) Six new species recorded for Grid area: Band-tailed Pigeon - 1 collected 27 July Long-eared Owl - 1 collected 28 July Whimbrel - 4 observed (26 July [3]; 29 July) Northern Phalarope - 16 observed (21, 23, 26 July) Heerman's Gull - 3 observed on 23 July Nighthawk sp. - 1 observed, 23 July 2.) High concentrations of Storm Petrels around Points "D," H," and "J." 3.) High concentrations of Cook Petrels in southern section. 4.) Overall preponderance of birds in eastern and southern sections. 5.) Occurrence of all dark (Leach's) Storm Petrels in regular numbers. Table 6 was prepared to test the hypothesis that the east-to-west abundance of birds on the Grid is dependent in part on the distance to land. Grid track observations were divided in half along the oblique line connecting points "Birch" and Oak. Notice that this diagonal very nearly parallels the coastline (see Figure #2). It was hoped that this epdata would give more interpretable results than the east-west division in respect to distance from land. The results were somewhat surprising. Instead of showing great changes from the standard east-west split, in- spection of Table 6 shows few discrepancies; the Jaeger data being the most notable. As a 45° rotation of the central dividing axis produced no marked changes in the density of the two halves, a further rotation of 45° was calculated. The overall densities are figured in Figure 3. Notice that by proper selection of bisectors, all areas except the 1/8 Grid section labeled "A" can be found to lie in that half which is con- sistently two times more dense than the opposing half. All is not entirely lost in this confusion if all three bisectors are super empirised on one another and tabulated as indicated in "Solution 'A' disting rela- tive densities in six portions of the Grid. "Solution 'B'" shows the densities computed for the routine six subsections (see Figure #2 ). Where numbers are high enough, this approach to variation of bird densi- ties within the Grid is fairly satisfactory. Even more graphically simple is "Solution 'C' phase where the sections are assigned plus and minus values relative to the overall density. Any additive unit that will serve to give a quick summary of direction and degree of non-randomness, Is applicable, as the measure is only relative. [In this case the formula is: the difference between the overall density and the subsection density X 100 roundoff, and attach appropriate sign.] When bird numbers are too low to be of statistical validity "Solution 'D'" may be most meaningful.