194 Proceedings of the Royal Society of Edinburgh. 
extracts from the Journal of Mr James Baillie Fraser, giving 
an account of his Journey to the Sources of the Jumna and the 
Ganges. At p. 1S7. of our present number, will be found an 
interesting notice of this journey. 
Jan. 4. 1819. — A paper by Dr Brewster w as read, ‘‘ On the 
action of uncrystallized surfaces upon light.” This paper con- 
tained an account of tw^o new classes of colours, formed by re* 
flection, and liitherto unnoticed. One of these classes w^as ca- 
pable of explanation by principles already known ; but the other 
had its origin in a new property of light, which promises to be 
of great utility in its practical applications. The author point- 
ed out its application to a new instrument for distinguishing 
precious stones, for discriminating mineral bodies, and for de- 
tecting adulteration in oils and other fluids. The instrument 
itself, constructed by Messrs P. & G. Dollond, was exhibited 
to the Society. 
Jdn. 18.— A paper by Dr Ferguson was read, On the 
poisonous fishes of the Carribbee Islands.” 
The author endeavoured to prove, that in all the larger Ashes 
of prey, the poisonous quality w as a rare and accidental occur- 
rence, and that it wras found to be present only at certain sea- 
sons of the year, in one or tw^o of the smaller species of fish, 
more particularly in the yellowy-billed sprat, (the sardine dore 
of the French, and Clupea thryssa of naturalists ;) from whence 
he inferred, that the larger voracious fishes, such ?-s the bara- 
costa, (Perea major of naturalists,) &c. became poisonous only 
at the times they had recently been preying upon the smaller 
poisonous prey. The notion of these being made poisonous, 
fi’om being found in copper banks, or their eating the stinging 
blubbers, (the Medusae and Hoiothuriae,) w^as refuted. In re- 
gard to tests, it was shewn, that none could be depended upon ^ 
That nothing whatever could be discovered from inspection of 
the fish ; that the boasted test of boiling a piece of silver 
wdth the suspected fish proved nothing, whatever might be its 
actual quality ; that so far from there being any marks of dis- 
ease in the viscera, or other parts of poisonous fishes, they 
were generally found to be in the best season, and of the highest 
quality, in all respects. 
The poison of the yellow-billed sprat, was supposed to be 
inherent in the animal at certain seasons of the year, and not 
