Four of these stalls are situated to the north, with a view of the sea, and the other four overlook the garden.  They are separated from each other by a simple partition, and all open on a wide central corridor that leads to the aquarium.  Before reaching the latter we find two offices that face each other, one of them for the lecturer and the other for the preparator.  These rooms, as far as their arrangement is concerned, are identical with the stalls of the workers.  The laboratory, then, is capable of receiving twenty-three workers at a time, and of offering them every facility for researches.

[Illustration:  FIG. 3.—­GENERAL VIEW OF THE ROSCOFF LABORATORY.]

The aquarium is an immense room, 98 ft. in length by 33 in width, glazed at the two sides.  It is at present occupied only by temporary tanks that are to be replaced before long by twenty large ones of 130 gallons capacity, and two oval basins of from 650 to 875 gallons capacity, constructed after the model of the one that is giving so good results at Banyuls.  At the extremity of the aquarium there is a store room containing trawls, nets of all kinds, and mops, for the capture of animals.  Here too is kept the rigging of the two laboratory boats, the Dentale and Laura.  Above the store room is located the director’s work room.

A wide terrace separates the aquarium from the pond.  This latter is 38 yards long by 35 wide.  Thanks to a system of sluice valves, it is filled during high tide, and the water is shut in at low tide, thus permitting of having a supply of living animals in boxes and baskets until the resources of the laboratory permit of a more improved arrangement.  This basin is shown in Fig. 3.  It is at the north side of the laboratory as seen from the beach.  Here too we see the aquarium, the garden, and a portion of the shore that serves as a post for the station boats.

We must not, in passing, fail to mention the extreme convenience that the proximity of the aquarium work room to the pond and sea offers to the student.

This entire collection of halls, constituting the scientific portion of the laboratory, occupies the ground floor.  The first and second stories are occupied by sleeping apartments, fourteen in number.  These, without being luxurious, are sufficiently comfortable, and offer the great advantage that they are very near the work rooms, thus permitting of observing, at leisure, and at any hour of the day or night, the animals under study.

Everything is absolutely free at the laboratory.  The work rooms, instruments, reagents, boats, dwelling apartments, etc., are put at the disposal of all with an equal liberality; and this absence of distinction between rich or poor, Frenchmen or foreigners, is the source of a charming cordiality and good will among the workers.

Shall we speak, too, of the richness of the Roscoff fauna?  This has become proverbial among zoologists, as can be attested by the 265 of them who have worked at the laboratory.  The very numerous and remarkable memoirs that have been prepared here are to be found recorded in the fourteen volumes of the Archives de Zoologie Experimentale founded by Mr. Lacaze Duthiers.