they have only to observe if, in cutting it up, there flows away a sort of white water, or rather a kind of thin matter, which is, in every case, a certain sign that the fish is in the diseased state of which I have spoken above. D. Arthur O'Neill, Marquis del Norte, has told me that he has seen experiments tried on dogs, and that all have confirmed the exactness of this criterion. The symptoms of poisoning by the Barracoota are, a general trembling, nausea, vomiting, and acute pains, particularly in the joints of the arms and the hands. Sometimes the symptoms succeed each other with such rapidity that it becomes extremely difficult to determine with precision the different periods of the disease.
"When death does not terminate the malady, which happily is the more ordinary case, the virus is sometimes seen to cause pathological phenomena altogether singular. The pains in the joints become stronger; the nails of the feet and hands gradually fall away; the hair also, which is of a nature analogous to the nails, ends by falling off. These phenomena have been observed in many individuals, sometimes continuing during a great number of years. A person has been mentioned to me, who suffered in this way more than twenty-five years.
"It is a remarkable fact that when the Barracoota has been salted, it never causes any accident. At St. Croix, for example, they are in the habit of eating it only the day after it has been salted. Does salt act as an antidote to the poison of this fish?
"I have not myself been a witness of any cases of poisoning by the Barracoota, and I have only
