[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

average number sighted over a number of seasons has tended to rise (see Table II), and for the decade 1944 to 1953 it averaged 173. During the 1954 season there was a spectacular increase to 495 sighted although the lookout equipment and personnel were essentially the same as had been used throughout the previous decade. Long spells of clear weather frequently made it possible to see a greater distance than usual, but there can be little doubt that humpbacks were considerably more abundant than usual in Cook Strait during 1954. This has been confirmed by a sighting total of 190 in Cook Strait during 1955.

A comparison of variations in the numbers and time of migration of humpbacks in the different seasons depends largely on the representativeness of those sighted as compared with the total humpbacks passing through Cook Strait each

season. Scanning is carried out from a lookout situated on a headland about 200 feet above sea level on the south-west side of the narrowest part of Cook Strait (i.e., 14 miles), and during average visibility spouts can be seen for a distance of at least nine miles. There is a permanent lookout man scanning the Straits from dawn till dusk and when not hunting, the six men who make up the chaser crews also take part. At the times that the catches are less than the factory can deal with, the towing vessel frequently leaves before daylight and sails up to 20 miles northwest by dawn to be in a position to sight whales which may have passed through the Strait during darkness. The whales which must pass unobserved are those which during darkness traverse the whole distance that can be scanned next morning from the lookout hill and sometimes the distance that can be scanned from the tug further north in addition. Although there are 14–15 hours of inadequate light for scanning, it would require several hours for a whale at the normal cruising speed of four to five knots to pass the area which can be seen next day. This reduces the effective cover to approximately nine hours or less in normal weather, but there are a number of days in each season where misty rain reduces visibility to almost zero, or gales confuse the surface of the sea to such an extent that spouts can be distinguished only when very close. Any estimate of the proportion which passes at these times or even travels unobserved in good weather would be highly speculative, but the height of the lookout, its situation at the narrowest part of the course into which all the local stream of whales have to be funnelled, and the number of experienced observers scanning Cook Strait make it probable that their sightings, although not large, represent a larger proportion of the local stock than would be seen from most whaling stations.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

For a number of seasons (1940–1953) the dates of sightings were not recorded, but a record of the numbers for each of the seasons from 1944 was carefully kept by notches cut into one of the lookout chairs. Counts of these have provided the sighting data for 1944–1953, and from Table II it can be seen that catches during that time represented an average of 61% of those sighted. The dates for catches were recorded for these seasons and also for 1940–1944. In lieu of dated sighting data these have been used to give the main sequence for the 1940 to 1953 seasons. During those seasons for which there are dated records of both sightings and catches it can be seen from Fig. 3 that there is good agreement between the two sets of data excepting for a slightly higher ratio of sightings to catches near the middle of the season.

The time sequence and relative density of humpbacks per week for 35 seasons has been summarised in Fig. 3. The earliest and latest record in this period were May 2 and August 29 respectively, thus covering a period of 17 weeks, but Fig. 3 shows that the greater part of the migration through Cook Strait occurs in June and July. Approximately three-quarters of the total were noted in the seven weeks period June 12–July 30, with the highest weekly sightings from June 19–July 9. Fig. 3 indicates a tendency for humpbacks to pass through Cook Strait in small numbers early in the season, followed by increasing numbers per week until the peak period between mid-June and early July, followed by a decline in numbers

per week until the end of the season. Part of the apparent rise and fall in records per week as indicated by the combined weekly totals is due to the inclusion of seasons with a late start or early termination, but the individual seasons also show a gradual rise and fall, although with varying dates for commencement and termination.

The number of humpbacks sighted (or caught) per week during each of 36 seasons is listed in Table 3, which shows a very considerable weekly variation in nearly all seasons, although the tendency towards larger numbers at mid-season is apparent in nearly all cases. A more useful criterion for considering seasonal variation in time sequences, is to compare the time required in different years for equivalent proportions of each season's total to pass through Cook Strait. The proportions used are as follows:—Time intervals in days for 12½%, 25%, 50%, 75%, 87½%, and 100% of those recorded. This arrangement results in a growth increment curve for the population each season, but for convenience is shown in column form in Fig. 4. The commencement of each season's column is arranged against the date for the beginning of the season, so Fig. 4 also shows the dates on which each increment was reached, as well as the length of time required for each increment The peak is taken as the date by which 50% of the sightings or catch had been recorded, and

Fig. 3.—The total number of humpbacks sighted or caught per week during 36 seasons at Tory Channel, Cook Strait (upper curve). The total number per week sighted during 14 seasons (middle curve) compared with the total number caught in the same seasons (lower curve).

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

the 25% on either side of the peak shows the time required for the central half of the migrating stock to pass through Cook Strait.

From Fig. 4 it can be seen that the individual seasons have varied in duration from 64 to a 110 days (averaging 86 days) and there is therefore no significant correlation between the start of the migration and its termination. The longest seasons, 1920 and 1937, commenced on May 2, which is earlier than usual but did not end until August 20, which is later than normal. In 1940 the season commenced more than four weeks later but ended a week earlier. There can, therefore, be no doubt that the whales sometimes travel in a more compact group which passes in a shorter period, while in some other seasons they are spread out as they migrate through Cook Strait and the period of frequent sightings may be prolonged. There is, therefore, no correlation between length of season and number of sightings or size of catch.

The date that the first sighting or catch occurred does not correlate with the peak of the season, which, as calculated above, varies through five and a-half weeks ranging from June 21 to July 29. As the date of appearance of the first whale may be followed by many days before the main group of whales commences to pass through the Strait, it is no consistent guide to the time that the main group of whales will pass through. In the most extreme case a whale was caught on May 20, in the 1934 season, but no more were seen during the next three weeks, and regular sightings and catching did not commence until June 11, although the weather had been suitable for sightings throughout that period.

Of all the intervals shown in Fig. 4 there is greatest variability in the quarter made up of the first and the last 12½%. It ranges from 16 days (in 1940) and 55 days (in 1949) to 69 days in 1920. In these cases the first 12½% was noted in four days in 1940, indicating a season where the main group of whales followed very closely after the first ones passing through the Strait, while 35 days elapsed in 1949 and 46 days in 1920 before the same proportion was observed. In most seasons the time taken for the last 12½% of the sightings tends to be shorter than the first 12½% because whaling ceases if the whales are becoming too widely spaced and there may be a few late migrants which pass through Cook Strait after the lookout has ceased. Even so there are six seasons in which the terminal 12½% of the records were spread over a substantially longer period than the first 12½%. If whales are widely spaced early in the season some of the whalers believed that this was a probable indication of a similar spacing at the end of the season. There are too many exceptions to this for it to be used as a guide to the distribution of whales later in a season. Seasons such as 1933 and 1943 commenced slowly but ended relatively quickly, while others, such as 1952 and 1953, commenced quickly, but Fig. 4 shows that the late whales were considerably more widely spaced.

It would be expected that the second 12½%, being closer to the peak of the migration, would travel in less time than the first 12½%, and this is the case in all except four seasons. The latter show the second interval of almost equal length to the first except for the 1940 season. In all seasons except 1917 the penultimate 12½% travelled through Cook Strait in less time than the terminal group.

The central half of the migrating stock is indicated as the two 25% groups on either side of the date for half the total sightings. There is closer agreement between the duration of each of these 25% groups per season than between any of the other intervals. In all except the 1915 and 1951 seasons, the interval occupied by the middle half of the migrating stock is substantially less than for the other half. The period ranges from 18 to 39 days, whereas that for all seasons combined is 32 days. In spite of the wide range of variation in the time taken by the various proportions of the migrating stock, the tendency for humpbacks to migrate in a group which is spaced out most widely at the van and rear and grouped more compactly in the middle is shown in all except two of the 35 seasons.

The central grouping as indicated by catch, is most spectacular in the 1942 and 1948 seasons, in each of which the middle half of the catch was taken in 18 days (see Fig. 4), whereas the total season lasted for 95 and 87 days respectively. The dates of sightings were not recorded for these two seasons. However, the ratio of catch to sightings is almost invariably lowered near the peak of migration because the factory is then working to capacity, and it is not possible to deal with the same proportion of those seen as can be done when the whales are more widely spaced. During these two seasons as well as the others between 1940 and 1953, the central half of the local humpback population probably traversed Cook Strait in a more compact group than indicated by the catch data alone. This was certainly the case in 1954, when 28 days were occupied in catching half the seasons's total, but as shown in Fig. 4 the central half of those sighted occurred in only 18 days. One hundred and twenty-five humpbacks, or more than a quarter of the total sighted, passed through Cook Strait in one week (June 26–July 2).

The two seasons, 1915 and 1951, which do not show the presence of a more compact central group, both had the highest weekly sightings early in the season, and the grouping into proportions of catch from the starting date as shown in Fig. 4 obscured the early concentrations in each case. It is clear, then, that all seasons have shown some period with a marked concentration of whales, and there has been no evidence of more than one such peak in a season.

In all seasons except 1915 and 1951 the maximum concentration has occurred near the mid-point of the central half of the sightings and therefore coincides closely with the date for the completion of the first half of each season's sightings. This date is calculated from the total seasonal data, and is dependent on the distribution of the whales throughout the season, whereas the commencing or closing dates for a season can be substantially altered by the presence or absence of a single whale. The date for the mid-point of sightings or catch will not be varied greatly unless the whole whale group is moving through Cook Strait earlier or later than usual. The mid-point has therefore been selected for comparing the times of humpback migration in different seasons. It shows a fluctuation through a maximum interval of 38 days between June 21 (1947 and 1950) and July 29 (1933), whereas the equivalent point for all seasons combined is July 5. Some of the smaller differences in the time of the apparent peak in different seasons are caused by variations in the length of adverse weather conditions which have prevented the sighting of some of the whales passing through Cook Strait. Such periods of unfavourable sighting conditions lengthen the period required to record a particular proportion of the catch, and result in an apparent shift for the date of the mid-point or peak of total sightings.

The factors determining visibility from the Tory Channel lookout are so frequently localised and subject to rapid change because of the unstable meteorological conditions in the Cook Strait area that it has not proved possible to use the weather records from the nearest meteorological station as a sound basis for correlating the distribution of sightings with weather. It is unlikely, however, that a range so large as 38 days in the time of the peak could be due solely to local weather conditions. It is more probably a real indication that the date of movement for the main group of humpbacks does vary considerably.

The cause is not apparent, but may well be due to annual differences in climate at the site of the beginning of the migration resulting in an earlier or later start, which is reflected in the variations noted at Cook Strait Records from the latter area illustrated in Fig. 4 show the seasons which have been early or late as compared with the average for 35 seasons, but there is no indication of the factors which may have affected the time of the migration movement in the different seasons. There are also no regular cyclic sequences of seasonal patterns recognisable from the available Cook Strait data. However, the fluctuations evident in Fig. 4 show some sequences