The well-known work of Hensen (1887) was the first serious attempt to determine the actual number of individual plankton organisms in sea-water. Hensen's method consisted in straining a vertical column of water through a net of fine-meshed bolting silk having 6,000 to 6,500 meshes per square centimetre, the average length of the side of a mesh being 50 µ. A carefully measured sample of the organisms retained by the net was taken and the number of organisms in it counted. From this, since the area of the mouth of the net and the distance through which it was drawn were known, the number of organisms in a unit volume (1 litre) of water could be calculated, with the help of a coefficient of filtration for the particular pattern of net employed, which was determined experimentally. Many small organisms, however, escaped through the meshes of the net, and Lohmann (1902, 1908) made a special study of these by filtering through hard filter paper or closely woven taffeta silk, by an examination of the filtering apparatus in the “houses” of Appendicularians, and finally, for the quantitative estimation of the smallest organisms of all, by subjecting samples of sea-water to the action of a centrifuge making 1,400 revolutions per minute for a period of 7 minutes. The use of the centrifuge has been continued by Gran (1915) and by Lebour (1917). For the full literature of the subject the reader is referred to Gran (1915) and Lohmann (1911). In the course of my work on the cultivation of plankton diatoms (1910, 1911) I became convinced that even the quantitative method based on the use of the centrifuge fell very far short of giving the total number of organisms actually present in a sample of sea-water, and the results recorded in the present paper show that this is certainly the case.

During the year 1918 several young fish brought in alive from the tow-nets were kept in small aërated aquaria at an even temperature and fed on plankton. In this way feeding habits were noted and some post-larval fish went through a metamorphosis into the adult form, notably Labrus bergylta the Ballan Wrasse, Solea vulgaris the Sole, Solea lascaris the Lemon Sole, and three species of Lepadogaster. Young Whiting and Pollack were also kept alive for some months and fed eagerly on most of the food that was given to them.

Investigation of the food of larval and post-larval fish from the townets has been continued throughout the year (1918) in the same way as previously, by examining the contents of stomach and intestine in the fresh state. From May 24th a new big net made of stramin, with a wooden frame 36 inches by 25 inches, was used in addition to the ordinary tow-nets. Any live specimens, and these were chiefly caught in the big net, were kept in small aërated aquaria standing in running water and fed with plankton in order to see, when possible, what food they ate. In this way a few post-larval fish went through the metamorphosis into the adult form. Details of these will be found in another paper in the same number of the Journal (page 9).

The tow-nettings were regularly examined by Miss Webb in order to see whether the fish were feeding on what was commonest at the moment, and this was found to be usually the case, so far as the Copepods were concerned, although most of the young fish seemed to select their food to a certain extent. In any month the Copepods which are at that time commonest in the plankton are those most eaten by the young fish, but most of them have preferences and select in various degrees. Thus in June Calanus finmarchicus and Acartia clausi were the commonest Copepods, Temora longicornis coming next, Pseudocalanus elongatus and Oithona similis not occurring so frequently. The commonest young fish in the tow-nets in the same month were the Whiting, Pollack, Rockling and Ballan Wrasse (Labrus bergylta).

Several Gobius species and also Aphya and Crystallogobius are common near Plymouth, from up the estuaries in the neighbourhood of Saltash to the west and Chelson Meadow to the east, as far as the open sea, well beyond Rame Head and the Eddystone Lighthouse. It has, however, always been difficult to determine the young of the various species as they usually differ very little from one another and it is hoped that the following notes may be a help in the elucidation of the subject.

Upogebia deltura Leach and Upogebia stellata (Mont.) are both of common occurrence on the coasts of Devonshire and Cornwall, but in the adult stage they are not easily obtained, owing to their burrowing habits. The larvæ, however, are very common in the tow-nets, occurring throughout the year with the exception of the midwinter months, and are particularly plentiful from June to August. The two species are closely allied, and for some time it was believed that the differences between them might be merely sexual, and not specific (Bell 1853). They are now clearly recognised as separate and distinct species, and De Morgan (1910) has obtained males and females of both forms. Upogebia deltura Leach may be distinguished by its greater size, sometimes five or six inches in length when full-grown, as well as by its broad abdomen with soft abdominal pleura. The two joints of the chela on the first thoracic leg are almost equal in length. Upogebia stellata (Mont.) is an altogether smaller and more slender animal, seldom exceeding two and a half inches in length. The abdomen is narrower, and the fixed finger of the chela very much shorter than the dactylus. There is also a small point on the side of the carapace, just over the base of the second antenna, which has been established as a specific character by De Morgan (1910). In the living state the body of the animal is covered with orange-red spots (hence its name stellata), while Upogebia deltura Leach is of a uniform dirty cream colour.

Whilst employed at the Royal Naval Hospital, Plymouth, I have been kindly granted facilities for the investigation of marine bacteria by Dr. E. J. Allen, Director of the Marine Biological Laboratories, and although the pressure of other duties has made progress extremely slow the experience obtained with different media will perhaps be of use to future workers.

Though it is easy to write down this heading, it is by no means so to define it. The simplest and best definition seems to be “A dipteron which spends a considerable portion of its existence in salt or brackish water,” but our present chaotic knowledge of the early life histories of shore diptera does not advance us far towards clearing up the doubtful period, and we are thrown back on the shore frequenting perfect insects, many of which are of wide distribution.

The above paper on the Plymouth gobies (p. 48) was finished and in print before the new work of C. G. J. Petersen (“Beretning til Land brugsministeriet fra den danske biologiske Station,” XXVI, 1919) came to hand. Although a second description and differentiation of the common forms is now unnecessary after his accurate and detailed account of them, yet as a corroboration of his identifications and as a record of the occurrence and distribution of the young gobies to be found at Plymouth (three of which are not described by Petersen), it seems advisable to keep the whole of my work intact, although in some measure it contains a repetition of the conclusions arrived at by him. I have therefore altered nothing, except adding the words “up to 14 mm.” with regard to Gobius pictus in the table, p. 79, and merely confine myself to a few notes in this appendix with regard to any slight differences between his specimens and those from Plymouth. The three species not described by Petersen, namely Gobius elongatus, Gobius Jeffreysii and Gobius paganellus are apparently, with the exception of Gobius Jeffreysii, not present at all in the Danish waters. On reading his conclusions as to the specific value of the number of vertebræ and fin rays I feel still more strongly that the form hitherto known as the deep water form of Gobius minutus, and which I have here distinguished as Gobius elongatus, is a true species, although closely related to Gobius minutus.