ON THE UPPER OLD RED SANDSTONE OF DURA DEN. 121 
part of the fish-bed was then worked. A few good specimens were 
obtained, and were on view when the locality was visited during one 
of the excursions arranged in connection with the Geological Section of 
the British Association Meeting at Dundee in 1912. 
Work was resumed by our Committee on May 5, 1913, and pro- 
ceeded more or less continuously to the end of August 19138. The 
pit, opened in 1912, having been partly refilled, had to be cleared 
again. As stated in the preliminary report issued last year, a definite 
plan was followed in the excayations. The fish-bearing zone was un- 
covered and removed in successive sections (fig. 3). 
The sandstone layer, rich in fish-remains, is restricted to a zone 
about two inches thick. It lies at an average depth of nine feet from 
the surface, and is overlain by about four feet of comparatively barren 
sandstone, capped by about four feet of loose superficial materials. 
It was decided to work the fish-bed in the direction in which the fish- 
remains appeared to be most abundant. As the operations extended 
towards the mill-lade in the area marked A’ in fig. 3, the sandstone did 
not yield fishes, as if the limit of the rich fish-bearing zone had been 
reached in that direction. The arrangement was then made to carry 
on the excavations towards the stream and just north of the face of the 
old workings. 
The finest specimens of fossil fishes and the largest number were 
obtained in the middle section (area marked B in fig. 8) and in the 
immediately adjoining parts of the other two sections (A and © in 
fig. 3). The greater part of the area marked C in fig. 3 proved to be 
somewhat disappointing, though one slab containing twenty specimens 
was found there and a fine example of Phyllolepis quite close to the 
stream. Good specimens, however, were scarce in section © outside 
the limit of the rich fish-bearing zone. In the north-east corner of 
it near the stream a sandstone layer with fragmentary fish-remains 
was traced for a short distance. 
It is worthy of note that large scales of H oloptychius were obtained 
in the sandstone three feet above the fish-bed, and that fish-scales in a 
fragmentary condition were found scattered throughout the sandstones 
above that zone. No fish-scales were detected below that horizon, 
although the excavations were continued downwards for nearly two 
feet beneath that zone. 
Fine examples of sun-cracks were seen in the sandstone at depths 
varying from two to four inches below the fish-bed, and, at one 
locality, one inch above that horizon. This feature is suggestive, and 
probably points to desiccation as a cause of the death of the fishes 
in a shoal at this locality. 
In all forty-two slabs of stone with well-preserved fish-remains were 
obtained. These were photographed by Mr. Dunlop, and the photo- 
graphs were sent to Dr. Smith Woodward for determination. About 
fifty fragmentary specimens were collected which were not photo- 
graphed. The whole collection has been stored in an adjoining mill 
under lock and key. 
The expenses connected with these detailed investigations have 
exceeded the British Association grant of 751. and the contribution of 
