The exhalant siphon of Xylophaga is greatly reduced and its opening is situated within the burrow. Faeces accumulate in the posterior end of the burrow in a compact mass.

The ctenidia and labial palps of Teredo, Bankia and Xylophaga are modified and reduced, unlike those of the Pholadidae. The ctenidia are flat and homorhabdic and possess no ciliary sorting mechanisms; they consist of the outer demibranch only.

The ctenidia of Xylophaga are composed of a direct lamella; those of Teredo and Bankia are divided into anterior and posterior portions connected by a branchial groove which is an extension of the marginal food groove of the gill. The number of filaments in the anterior portion of the gill varies according to the species, and is of assistance in identification.

The labial palps of Teredo norvegica are least reduced; in Xylophaga dorsalts reduction is greater, though selection of food material is still exercised. The greatest reduction has occurred in Teredo megotara and in T. navalis where there is no selection.

The ciliated mantle tracts in Teredo and in Xylophaga play an important part in limiting the quantity of food particles passed from the ctenidia to the mouth.

The morphology of a complete series of developmental stages of Lernaeocera branchialis is described. Seven stages are passed through before reaching the gadoid (final) host.

The structure of the mouth tube and appendages is discussed in detail and compared with those of related genera in which the resemblances are close.

The free single nauplius and copepodid stages are immediately followed by four chalimus stages on Pleuronectes flesus; these show some dedifferentiation though segmentation is not lost. An explanation is given for the peculiar suspensory mechanism of the chalimus.

A second dedifferentiation follows the second well-developed pelagic phase in which the adult form is attained and copulation takes place. Some irregularity is noted in the details of the retrogressive metamorphosis of the maturing female on the gadoid host.

The rate of increase in body length in Lernaeocera decreases regularly up to the assumption of the adult form (stage VII). In this respect it is similar to the free-living copepods Diaptomus, Eurytemora and Calanus, but is in strong contrast to Caligus centrodonti which has an increasing growth rate up to the fourth chalimus stage. In the latter at this stage there is a marked acceleration in the growth rate unaccompanied by any change in shape or any other discontinuity, whereas in Lernaeocera there are early discontinuities in development, yet the growth rate is continuous when expressed graphically: it is a straight line rather than a logarithmic curve as would be expected from our knowledge of the higher Crustacea.

The size of the female is greater than that of the male: that of the female overlapping the male of the succeeding stage. Variations in size of the larvae in a stage group are not seasonal, and an explanation is offered for giant larvae.

Similar lernaeocerid larvae and males to those of Lernaeocera branchialis on Pleuronectes flesus were found on Solea solea and Cyclopterus lumpus. No characters of generic or specific significance could be found which would serve o t separate them from Lernaeocera branchialis so that their identity is not determined.

The late Dr E. J. Allen was the son of the Rev. Richard Allen, a Wesleyan minister and as such liable to change the sphere of his work every few years. For this reason E. J. Allen passed his childhood in Cornwall, went to school in Exeter, and lived during his late boyhood in the Worth Valley, going into Leeds to attend the Yorkshire College (now University of Leeds); his education was continued by the University of Berlin, an interval of teaching in Antigua, and University College, London, followed by a Royal Society research studentship at the Plymouth Laboratory. Excepting Antigua, every one of these epochs claimed in him its own loyalties: a few years ago he seriously considered going west to Cornwall to end his life where he first remembered it; he much admired Franz Eilhard Schulze, who taught him at Berlin, and W. F. R. Weldon, his Professor of Zoology at University College, yet kept a lively gratitude for what he had learned from Professor Miall at Leeds.

To this very varying environment during the first third of his life we may attribute through the next half-century much of his universal power of sympathy and contact with all sorts and conditions of men. Unselfish by nature, he had the example of a singularly wide-minded father; then bred as a boy in Cornwall, Devon, and West Riding, as a man among London University students and Berlin University students, with the West Indian interlude— there was no class or nation who could not feel an understanding with Allen.

1. The number of species found in this investigation is six, of which four are marine and two brackish water.

2. The mean value of the body weight and the fecundity varies with the species. The size of a species bears no relation whatever to the reproductive capacity.

3. The fecundity is subject to seasonal variation, being generally higher in the spring than in the winter months.

4. In general, a positive correlation exists between body weight and fecundity of the individuals of each species. This rule does not, however, apply to Gammarus duebeni, in which the fecundity appears to fall after reaching its maximum at the 19 mg. body-weight class.

An account is given of the distribution, life history and trematode parasites of Littorina neritoides on the Plymouth Breakwater; a few observations from other localities are included.

The smaller snails are most abundant on the exposed southern slopes of the Breakwater. Many of the larger ones live more or less permanently in water in small cylindrical pits on the top of the Breakwater. This habitat is apparently very similar to that described by Fischer-Piette (1932) as occupied permanently by this species at Cap Martin; Lebour (1935) has sometimes found it in water in other places in the Plymouth area.

The conditions necessary for metamorphosis of the larvae, the need of the adults for shelter from the force of the waves, and the requirements for spawning appear to be of more importance in determining the choice of habitat than the negatively geotactic and varying phototactic responses found in this species by Fränkel (1927).

The breeding season lasts from September to April. The males are ripe about two months before most of the females. Experimental evidence is given which makes it most improbable that there is any downward migration for spawning. From the examination of plankton samples it appears that there is a fortnightly spawning rhythm coincident with high tides, and that even the snails living in water discharge their eggs only at these periods.

The distribution of the size groups is discussed. The proportion of males decreases significantly in the larger size groups and it is probable that there is a difference in the growth-rate of the sexes.

Onchidella celtica lives in communities in intertidal rock crevices, emerging only when uncovered by the tide during the milder months (March- November). It feeds on young algae and diatoms, rasping them from the rock with the radula which is lubricated by a mucous saliva. The distensible crop and stomach afford ample storage space for the food. The stomach consists of three chambers (Fig. 2), and the dorsal (d) and ventral (v) chitinous plates of the posterior one, the gizzard, crush the contents and help to direct the soluble portion towards the ducts of the digestive gland. There are three of these the two anterior (Id, rd) are large and lead from the initial chamber of the stomach (ic), the third (pd) is small and joins the posterior wall of the gizzard (pzv). The digestive gland secretes into the stomach; no cellulase occurs. Feeding experiments suggest that the digestive cells of the gland absorb only fluid. Lime and excretory cells occur in the crypts of the digestive epithelium, the latter concerned with taking up excretory matter from the blood and elaborating it into spherical masses for evacuation with the faeces

The rapid growth of bacteria when sea water is stored in glass vessels, and the possibility that this is brought about by the concentration of food by adsorption on the solid surface, are discussed. Some evidence is presented bearing upon such adsorption.

The regeneration of phosphate from dissolved organic phosphorus compounds in the sea and in stored water is considered.

The American whelk tingle, Urosalpinx cinerea (Say), a gastropod enemy of oysters, has been introduced to Britain with American oysters and is now established in the rivers Blackwater, Colne, Crouch and Roach in Essex and on the Kentish Flats off Whitstable. Evidence available concerning the migratory powers of Urosalpinx is conflicting, but it is probable that migration is of local significance only in the spread of this pest. To infest new areas tingles or their spawn must be transported by human or other agency, for Urosalpinx has no free-swimming larval stage, the eggs being deposited in capsules from which fully formed young tingles emerge.

Females of Urosalpinx grow more quickly than males and reach a larger size. In males the annual increase in shell height never exceeds 0·2 cm. after the first 3 years, and is generally rather less. In females an annual increment of 0·2 cm., or slightly more, is general up to the age of about 7 years, after which it gradually falls. Males and females of the same age differ in height by about 0·25 cm. Growth marks are frequently to be seen on the shell and are of value in assessing growth rates. In both sexes a maximum age of 13–14 years may be reached. Urosalpinx reaches a much greater average size in Britain than in its natural habitat on the Atlantic coast of the U.S.A.

Females may deposit some spawn during the first 2 years of life, but the amount is insignificant in relation to the production of older tingles. Spawning begins when the water temperature in its seasonal rise reaches 12° C. Adult females may deposit an average of 25 egg capsules at a single laying, but it is possible that further capsules are deposited later in the season. The bulk of the spawn is deposited in May and June on British beds, but a few freshly laid capsules may sometimes be found in August and September. The average period of incubation is about eight weeks. Young tingles usually begin to emerge early in July. The average number of young emerging from each capsule is 11–74. Urosalpinx appears to be more prolific on British oyster beds than in America.

In a previous paper (Smith, 1940b) the chemical changes observed in ocean sea water flowing across the shallow banks on the west coast of Andros Island (Bahamas) were reported. High salinities were produced by evaporation and calcium carbonate was precipitated. From the data obtained under natural conditions a maximum value of the solubility product constant of calcium carbonate in sea water was suggested. There was no reason to believe, however, that this value of the constant represented the true equilibrium conditions, and experiments have since been made in the laboratory with a view to bringing water from these banks into equilibrium with solid calcium carbonate.

If the number of infections with (a) trematode parthenitae and cercariae using Littorina neritoides as first intermediate host only, and (b) encysted metacercariae using L. neritoides as second intermediate host only, are plotted against the size of the snails, two different curves result. The first shows a low rate of infection in the small size groups, but a steep upward slope rising to 91% in the large size groups. The second shows a curve increasing uniformly to 87% infection.

Possible interpretations are discussed, and it is concluded that the difference is probably due to the fact that primary infections cause accelerated growth in the host.

Although the larva of Porcellana is one of the earliest known decapod larvae and at a later date the different stages of P. longicornis were well described by Sars (1889), there is yet much that is unknown. Few larvae have been actually hatched from the egg and in no case has the number of larval stages been definitely ascertained. In order to fill in some of the gaps in our knowledge a detailed study of the two common Plymouth species was begun (P. longicornis and P. platycheles). Unfortunately, war conditions made it difficult to obtain material and many specimens were lost as the result of air-raids; it was therefore not until the summer of 1942 that the work could be finished.

It is now possible to distinguish between the two species in all stages. Both have been hatched from the egg, larvae from the plankton have moulted and the post-larvae have been obtained from the last larvae. A comparison with Gurney's (1938a) notes on the species from the Red Sea and his general comments on the larvae of the genus and its relatives (1942) is of much use. These, with the first larvae of Petrolisthes armatus hatched by myself at Bermuda and certain larvae described by other workers, have made it possible to make some progress in the elucidation of the relationships of the larvae.

Porcellana longicornis and P. platycheles are both common at Plymouth; but whereas the latter is confined to coastal waters and lives between tidemarks and down to about 3 fathoms, the former has a much wider range and although occurring withP. platycheles between tidemarks extends to well beyond the Eddystone grounds.

The faunal analysis of forty-eight samples taken with a coarse tow-net across the intertidal waters of Kames Bay during 1936 shows that the night migrant crustacean population falls into two well-marked groups: (a) those species which live in the intertidal sand and perform simple vertical migrations at night-time into the intertidal waters, and (b) those which are carried in by the tide into the intertidal area from various habitats which lie beyond the low-water mark of spring tides. In the former group the species of the genera Bathyporeia and Pontocrates are dominant and they retain their zonation as migrants, the population consisting of individuals in all stages of maturity. This is in sharp contrast to the latter group, typified by Gammarus locusta and Idotea viridis, which occur over the whole tidal range with a population in which the young immature stages are dominant. The Haustoriids Urothoë brevicornis and Haustorius arenarius occur as inhabitants of the sand, the former appearing occasionally in the tow-net samples and the latter absent from them. The Cirolanid Eurydice pulchra performs both day and night migration for the purposes of feeding.

Discussion of the results indicates the lines along which further investigations may be carried out.

It is shown that the average number of embryos produced by a gravid oyster is approximately doubled each year until the oyster is 4 years old, when it may be regarded as adult. Estimates have been made of the quantities of embryos produced by one 1-year-old, eleven 2-year-old, fourteen 3-year-old and twenty-two adult oysters; the mean values for these year-classes are 91,600, 218,100, 462,600 and 902,900 respectively.

Since it is probable that the proportion of a population of oysters which matures in the female phase increases progressively up to the age of 4 years, it follows that adult oysters are for this reason superior to 2- or 3-year-old oysters as breeding stock, quite apart from the increased number of embryos produced by each oyster.

1. In order that the Harvey method shall estimate without avoidable error the concentration of marine or freshwater phytoplankton, it is essential, firstly

The distribution of Arachnactis albida in the Celtic Sea in 1937–9 is described. The species appears, to have a definite centre of distribution in the neighbourhood of lat. 50° N., long. 10° W. The larvae have been taken as far south as lat. 48°, and eastwards up to the mouth of the English Channel, but they are absent from the tongue of low salinity water continuous with the St George's and Bristol Channels. Their distribution in the remaining area can be correlated with the cyclonic circulation.

The larvae obtained from the Celtic Sea are specifically identical with those found off the west coast of Ireland and with the typical A. albida of the Faroe-Hebrides region.

An attempt has been made to deal briefly but quantitatively with all size groups of the fauna and flora in a marine mud deposit.

It has been necessary to propose the following new terminology: the Macrobenthos, which is here equivalent to the macrofauna, the Meiobenthos, under which term are included copepods, nematodes, foraminifera, etc., and the Microbenthos, comprising the rest of the protozoa, bacteria, bottom diatoms and other algae; planktonic diatoms and coloured flagellates also occur on the bottom in the region investigated but are not regarded as true microbenthos.

A quantitative ecological study of the microbenthos has been started and the habitat is first described, stress being laid on points of importance to the microorganisms.

The census method for bacteria was an agar-plate method and that for the protozoa and diatoms was a dilution culture method modified from that used by soil microbiologists.

These methods give minimal values for the total population and may safely be used for comparative purposes. The results are still tentative.

Typical figures for the minimal numbers and volumes of living protoplasm per gram of dry mud in the top ½ cm. layer are given in Tables I-IV.

The surface layer is much richer in all types of organisms than is the mud deeper in the cores. There is considerable local horizontal variation in numbers of bacteria and protozoa, suggesting dense aggregations.

Diatoms contribute by far the greatest bulk to the total volume of living micro-organisms; planktonic diatoms, particularly during the summer, completely outweigh the true microbenthos.

The Council and the Officers. During the year the Association has suffered a heavy loss in the death of Dr Guy Wood, who had been a Governor representing the Fishmongers' Company since 1931 and had been Honorary Treasurer for the past seven years. The Council also regret to note the death of Dr Cresswell Shearer, F.R.S., a Founder of the Association.

Major E. G. Christie-Miller has been nominated by the Fishmongers' Company to succeed Dr Guy Wood as a Governor, and at the meeting held in October he was elected Honorary Treasurer

Four ordinary meetings of the Council were held during the year. Three of these were held in the rooms of the Royal Society and the thanks of the Association are due to the Society for this privilege. One meeting, in April, was held at Plymouth: it took place in the Grand Hotel since there were at the time unexploded bombs in and near the laboratory premises. The average attendance at the four meetings was twelve.

Air Raid Damage to the Plymouth Laboratory. No additional damage of any consequence has been incurred during the year, though a few windows have been broken. The delayed action bombs which fell shortly before the April meeting were safely removed.

For the first three months of the year the staff of the Laboratory was largely engaged in repairing the damage occasioned in the March raids and though much will need to be done before the buildings are fully restored most of the laboratories have been made usable.

1. Measurements of Carcinus, comparable to those taken by Weldon at Plymouth 40 years ago, have been made on material from three Irish localities, the observations in each area extending over three years. The three localities afford different environments, one having no silt, one a moderate amount and the other much silt.

2. The results support the view that the change in the ratio frontal width/ carapace length with increase in body size is due to differential growth in the individual and is not caused, as Weldon supposed, by the continuous removal through natural selection of those crabs with a relatively wide frontal aperture. Measurement on the growth of individuals confirms this view.

3. A striking fact which emerged from Weldon's work was that the ratio mentioned above showed a successive diminution in each of the three years covered by the observations. A precisely similar diminution has been found at each of the Irish localities, and it is shown that the annual differences are mathematically significant.

4. The results disprove Weldon's hypothesis that the change in the ratio is correlated with the slow accumulation of silt in Plymouth Sound. The Irish locality with most silt has the widest frontal aperture and the intermediate locality the narrowest. It is also shown that the yearly trend towards a lower value for the ratio cannot be continuous, for the rate of change is too rapid to be maintained indefinitely, and the values obtained in the Irish localities overlap those at Plymouth 40 years earlier. Possible explanations of the changes are discussed.

5. The differences between populations from the three localities in any year are much less marked than the annual differences at one locality, and are not generally significant. Though small, however, they do show a consistent sequence among the three localities (but not corresponding to the order for siltiness).

Manila ropes and trawl twines, and cotton nets, have been immersed in clean sea water for periods up to 14 months, some with alternate wetting and drying, some wet all the time, but in both cases accompanied by storage in air after immersion. All the untreated specimens perished, but the 700 lb. manila ropes retained at least 40% of the original tensile strength.in every case in which a copper soap had been included in the treatment. A previous treatment with cutch is a considerable advantage as shown by the 500 lb. manila ropes and trawl twines, though cutch alone is worthless. Coal tar distillates after cutch are, however, more useful than when used alone, and their mixture with copper soaps gives excellent results. Cutch fixed with bichromate gives tolerably good preservation on nets up to 6 months, but is a poor preservative for trawl twines and ropes which are expected to have a long life. The choice of the copper soap used depends upon the material to be preserved. Naphthenate is the most soluble and penetrates thick ropes excellently; on fine nets it leaches out easily. Oleate is cheaper, but its penetration is less; it does very well with tar on twines and nets. Resinate is the least soluble, but is useful with creosote oil or tar. On account of their greasy nature naphthenate and oleate act as lubricants in ropes.

Immersion tests on other preservatives, kindly suggested by Dr S. G. Barker, are to be begun.