The following paragraphs give a brief comparison of our findings with two recent test experiments on fouling and growth, one from the Northern Hemisphere page 14 and one from the Southern. Corlett (1948) describes observations in the Mersey Estuary during 1946 and 1947, using tiles and scallop shells as test units. It should be noted that surface, texture, size, and the time period for immersion differed in our and Corlett's experiments, but the same groups of organisms set in both. However, the most conspicuous groups in Corlett's investigation were not the same as in the present instance. Hydroids, barnacles, and mussels were the most important settling organisms with Corlett, while species of polyzoa, ascidians, and hydroids in that order were dominant organisms in Port Nicholson. The hydroids are important in both investigations. Otherwise, the groups that were most important in the Mersey Estuary were some of the least important in the present experiment. In most groups, a greater number of different species settled in the Northern Hemisphere—e.g., five species of barnacle as against a single species in Port Nicholson. This was also the case with the Australian investigation when compared with our New Zealand tests, and may probably be explained by the fact that in the present experiment the test blocks were confined to one situation of limited area, whereas Corlett, and Allen and Ferguson Wood, had several stations with varying conditions and at points some distance from each other. Species of the genera Tubularia, Mytilus, and Enteromorpha showed in Corlett's and our tests, while the ascidian B. schlosseri set at approximately the same season in both hemispheres—namely, early autumn in England and late autumn in New Zealand B. schlosseri was not, however, as prominent a species in Corlett's experiment as it was in ours.

Tests carried out in Australia by Allen and Ferguson Wood (1950) are more comparable with our experiment than are Corlett's, as approximately the same time elapsed (30 and 28 days respectively) between the raising of each test block. Also, Allen and Ferguson Wood had long-term units down for three and six months to obtain succession and rates of growth, and short-term units down for seven to fourteen days. The Australian workers, however, used glass plates for test units, thereby obtaining a more rapid check on the settling organisms than could be obtained from the wooden blocks. Their test units were suspended one to two feet below low-tide level, as against our four feet. The following genera, Ectocarpus, Enteromorpha, Ceramium, Mytilus, Sycon, Spirorbis, and the species, Obelia australis, Bugula neritina, and Galeolaria hystrix, were present in both experiments. None of the species of ascidians recorded by Allen and Ferguson Wood was found in the present study.

Marine Borers

As far as can be ascertained, the only published work on marine borers in New Zealand waters is a paper by Chilton in 1919. This is a general survey dealing with the systematics and habits of crustacean borers and the damage to wharves in Auckland, Lyttelton, Napier, and Wairoa. So far, in Port Nicholson waters, we have noted four species of borer. Two of them, a nereid, Nereis kerguelensis, and page 15 the Amphipod Chelura terebrans, seem to cause little damage. The burrows of Chelura terebrans are, however, very similar to those of Limnoria quadripunctata Holthius. It is possible that we have underestimated its damage, and that a greater amount of surface crumbling on some of the test blocks is caused by Chelura than is immediately apparent.

The third species of borer is the Isopod, commonly called the "gribble." At the request of Mr. Robert Menzies, of the Pacific Marine Station, California, specimens were sent to him. They were identified by him as Limnoria quadripunctata, a species described as new by Holthius in 1949, from Holland, and subsequently reported from Cape Town, Plymouth, and the Central Californian Coast. It had been assumed up to the time of Mr. Menzies' communication that the species present in Port Nicholson was Limnoria lignorum Rathke, as this species has been recorded from Auckland, Lyttelton, and Akaroa. Doubt has now arisen as to the correctness of Chilton's diagnosis of specimens from these three latter harbours. Since it is difficult to separate exactly the burrows of Chelura terebrans and Limnoria quadripunctata, we have referred in general terms to the damage done by these species as "Limnoria damage." Limnoria is undoubtedly responsible for the greatest visible destruction of the test blocks. The fourth species of borer found in the test blocks is the mollusc Bankia australis, commonly called the "ship worm," or "teredo."

Limnoria attacks the surface of the wood. As this becomes riddled with burrows, the surface disintegrates, and an accurate estimate of the attack can readily be made from month to month. Bankia, on the other hand, burrows below the surface, the burrows at first running with the grain, but later, as infestation becomes heavier, the burrows twist and turn in all directions. The only indication of "ship worm" in the wood is the presence of numerous small round respiratory holes on the surface. Although a count of these holes gives a fairly true measure of the numbers present, it gives no indication of the overall damage done to the interior of the wood. To obtain a more accurate picture of this damage and to obtain growth measurements, X-ray photographs were taken of the test blocks.

Infestation by Crustacean Borers (Fig. 2)

Limnoria numbers were estimated by ruling the test blocks into one-inch squares and counting the number of burrows. These usually follow the grain, and are marked at the surface by a series of respiratory holes. A strong needle run along the line of these respiratory holes will generally expose the tunnel for its full length. The male and female live together at the head of the tunnel, female first, and the young are liberated fully developed from her brood pouch. They commence to burrow on their own account away from the parent tunnel, so that badly infested timber will show branching tunnel systems. As the tunnels eventually become three page 16 or four deep, or even more, it becomes impossible to do more than make a very approximate estimate of the number present. This point was reached in the present experiment in about nine months. The short-term blocks rarely showed "Limnoria damage," indicating that more than a month's immersion under local conditions on this particular type of test block is necessary before much damage is apparent. Counts on the long-term series show accelerated breeding in October, and again in March, 1950. In both instances, the Limnoria population trebled the previous month's total. The first sharp rise in population numbers coincided with rapid temperature rise in spring. The cumulative effect of warm summer water probably accounts for the second increase in autumn. A decrease in numbers in November is shown, and must be due to some unknown factor affecting breeding peculiar to that block. Apart from this one instance, population numbers increased throughout the experiment. The vertical block and the rougher surfaces were consistently more heavily attacked than the horizontal block. The presence of a fairly thick coat of mud on the upper surface of the horizontal block probably accounts for this, since it was observed that where mud had accumulated there was little "Limnoria damage."

Infestation by Bankia avstralis Caiman (Figs. 3 and 4)

X-ray photographs of the first short-term block proved negative, and the first indication of Bankia attack was given by the second-month long-term block, where X-ray photographs revealed a light infestation of young animals ranging in size from 0·3 cm. to 1·8 cm. in length. By a careful comparison of the photograph and the surface of the test block, it was possible to see the minute respiratory holes of these young animals. In all cases, X-ray photographs failed to reveal the presence of Bankia in the short-term blocks. Isham, Moore, and Smith (1951) also record that "in panels exposed for only one month ship worms were neither large enough or numerous enough to furnish material adequate for growth rate determination." However, in February, when the long-term blocks clearly indicated a heavy set, it was possible in the present instance, with the aid of a hand lens, to make a fairly accurate count of respiratory holes on the short-term blocks.

Bankia respiratory holes are readily distinguishable from those of Limnoria, and counts of the holes give a reasonably accurate estimate of numbers present after the first few months. Counts from X-ray photographs give a check on the numbers obtained from counts of the surface holes. In large specimens of Bankia australis, the pallets often protrude slightly from the surface holes, and counting is then easy. The burrows are lined with lime, and usually go a short distance straight into the wood before turning to run with the grain. The fact that the shell valves and pallets are also mainly composed of lime greatly facilitates the count, as these calcareous structures show clearly in X-ray photographs. Accurate measurements of burrow lengths cannot be obtained after about the tenth month because of the number of intimately intertwining burrows.

Fig. 2.—Monthly incidence of Limnoria quadripunctala Holthius in the long-term test blocks. Temperatures shown by triangles. Full explanation of diagrammatic presentation of monthly temperature records on page 3 of text. *Approximate numbers.

From the second month (May, 1949), Bankia was clearly indicated on the X-ray plates. The burrow length for this month was from 0·3 cm. to 1·8 cm. Pallets could not be distinguished in specimens 0·3 cm. in length, were faintly visible in the 0·6 cm. to 0·8 cm., but were distinct in all specimens above this size. The smallest Bankia australis recorded on the X-ray plates was 0·3 cm. in length. There appeared to be no significant difference in growth rate between animals in the horizontal and vertical blocks. Over the first nine-month period (April to November, 1949), there were constant small reinvasions of the blocks by larval

Fig. 3.—Monthly incidence of Bankia australis Caiman in the long-term test blocks. Water temperatures shown by triangles.

stages. Numbers settling increased markedly from December, 1949, to April, 1950. The peak was February, when approximately 2,00 settled, with another 750 already established.

Figure 4 gives the growth of B, australis based on measurement of burrow lengths; the principal curve gives the length of the longest burrow in the block for each month. The other data for each month shows the groupings of the shorter burrows as a vertical line, with the average in each group marked by a filled-in circle, These latter groups are here considered as representing monthly sets, page 19 although—e.g., November—there are ten such groups which exceed the number of months and indicate more than one heavy set in some months. The principal curve, in general, is that of an ordinary growth curve, but shows a deceleration of growth rate in September and again in November (i.e., winter and spring). This is also noted by Isham, Moore, and Smith (1951), who have interpreted this as consequent from overcrowding, but in terms of the actual set found on the short-term blocks, the total population is still short of its maximum. In any case, the burrows measured are those of animals which have lived through the period of sharply increasing growth-rate (August and October), and did not benefit to the same degree as did those measured in the September and November blocks. This deviation of the curve accordingly must be considered as due to individual variation and not having the significance of the variation in Isham, Moore, and Smith's curve.

Fig. 4.—Growth curve of Bankia australis Caiman based on burrow lengths. Principal curve gives lengths of longest burrows. The symbol shows groupings of shorter burrows. Burrows measured at the end of each month.

results are a true indication of the growth and incidence of these borers in Port Nicholson. The sudden rise in numbers in February, followed by a marked decrease in March, is difficult to explain. A steady rise in numbers is more to be expected, with perhaps a flattening-off of the curve of total numbers about March with the decrease in temperature. The March and April, 1950, blocks were exposed to the same heavy settling as the February block as well as later settlings. A possible The incidence of Bankia australis over the thirteen-month period shows one or two unusual features which seem most reasonably to be accounted for by fluctuations in local conditions. Further study is necessary to see if these preliminary explanation seems to be that the adult Bankia in the February block liberated large quantities of spawn, and the resulting larvae settled mainly on that block. There has obviously been a decline in the survival rate, and for some reason spawn liberated in March and April failed to survive in high numbers on these blocks. Calcareous encrustations such as barnacles, oysters, the tubes of marine worms, and encrusting bryozoa form an armour against the entry of borers into wood, especially when their establishment precedes larval lodgement (Watson et at., 1936). This may have been a factor which prevented survival from large settlings on the March and April blocks, as on these other species were well established. However, these species were present also on the February block, and it is hard to see why they failed to reduce the numbers of larvae settling in February if they were one of the chief factors responsible for reduced settlement in March and April.

Several factors are known to influence "ship worm" attack, and different factors are thought to influence different species of "ship worm." Caiman (1919) claimed that pollution or muddy water rendered timber less likely to attack. These factors were present to a minor degree only, and seem unlikely to have influenced the numbers of settling larvae in the present experiment. The greater number of animals in the horizontal unit of the test blocks is in accordance with the findings of Shillinglaw and Moore (1947), who report that non-resistant cross and diagonal bracing below mid-tide level are very susceptible to attack. As all the blocks were placed at approximately the same depth, no indication of the vertical working range of Bankia australis can be given at the present time. It is generally agreed that, where the substratum is suitable, certain types of fouling organism are indicators of likely areas of borer attack. These include barnacles, hydroids, and mussels (San Francisco Bay Marine Piling Survey, 1923; Watson et al., 1936). Where these animals are already present in an area, borers have usually made their presence obvious also. Species of all the indicator groups enumerated above are present on the wharf piles in the vicinity of the present experiment; and, as indicated by the above description, borers have also made their presence obvious. In an area where barnacles, hydroids, and mussels suddenly make their appearance and where no borer troubles have been previously experienced, then one may page 21 expect borers, since salinities and temperatures that suit borers are usually within the ranges of tolerance of these other animals. The sudden appearance of barnacles, hydroids, etc., suggests a change in local salinities and temperature conditions which will suit the borers. Environmental conditions vary within the harbour area, and further experiments are necessary to give a more complete picture of borer activity in Port Nicholson. Most species of Bankia show preference for water of high salinity (Watson et at., 1936), and one would predict that B. australis may be less active on the eastern side of the harbour near the Petone foreshore, as the movement of the Hutt River water in the harbour on this shore must lower the salinity of the water.