5°o 



SCIENTIFIC NEWS. 



[Nov. 1 6, 1888. 



arrangement. As the fresh-water fauna is still im- 

 perfectly known, each of the tanks on the west side 

 may be fed either with fresh or salt water. 



The outer tank of about eight cubic metres' capacity 

 may serve, if needful, for animals of large size. 



The reserves are entrusted to a small floating viva- 

 rium. 



The laboratories of the first floor contain a long, flat 

 sink, which are provided with two gas jets for heating a 

 stove and a chafing-dish ; two cocks for fresh and salt 

 water, both at sufficient pressure to work filter-pumps ; 

 and two porcelain cocks for distilled water or alcohol. 

 These sinks are fitted to receive small aquariums of from 

 60 to 80 litres, which a cart may take to the edge of the 

 quay, and which may be lowered into the boat by means 

 of a crane. The most delicate animals may thus be con- 

 veyed from the water's edge to the tanks in the station. 



In each laboratory there is also a third gas jet for 

 apparatus which may be needed on the work-table. 

 Arrangements are being made to forward specimens and 

 preparations to such naturalists as may require them for 

 their studies. 



- — ♦-^t^i'^Mf-* — ■ 



ROMAN REMAINS AT LLANTWIT- 

 MAJOR. 



(Continued from p. 478.,) 



TN spite of the temptation to digress at this point, the 

 -*- plan of excavation was adhered to, and the trench was 

 continued. A large fragment of a pillar was found, then 



Fig. 2.— View of Trench. 

 the angle of the other side of the last wall, and a con- 



another corner was observed, and in this, the third en- 

 closed space, another fragment of a pillar, with orna- 

 mental base and many fragments of plaster, was 

 encountered. In this way the trench work was con- 

 tinued, opening up two more enclosed spaces and 

 passages, and yielding coins and a finial, among other 

 articles and fragments of interest. At a distance of 

 54 feet from the third enclosed space a hypocaust was 

 exposed for a length of 25 feet in the direction of the 



Fig. 3.— Finial. 



trench ; all the blocks of masonry and flues were very 

 distinct, whilst at the entrance, on some steps, the 

 skeleton of an old woman was found. 



Our illustration (fig. 2) represents portion of the trench- 

 work, showing the finial in situ, as discovered. Other 

 points, such as the hypocaust at the end, the angles of 

 walls, the bit of tesselated pavement, etc., will be 

 readily distinguished. 



The separate illustration (fig. 3) of the finial shows 

 the ornamental character of this architectural fragment, 



tinuation of the longer wall from 27 to 44 feet, where \ and there can be very little doubt but that it formed 



