Page:Sawdust & Spangles.djvu/288

From Wikisource
Jump to: navigation, search
This page has been validated.
250
SAWDUST AND SPANGLES

ceeded in persuading me to interest myself in their methods. I soon learned that these three-tailed fishes were the result of the Japanese system of breeding, of which they alone knew the secret, and when, on investigation, I learned that their waters contain many varieties of fish of gorgeous colors, I determined to spare no expense to possess a collection from this coast, especially after I learned that even Nature itself seemed reversed there, and that there are fishes in those waters that swim on their backs.

Supplying a trusty agent with the necessary money, I first sent him to Yokohama, with letters of introduction to some friends of mine. Here, assisted by the natives, he commenced forming his collection. The captured fish were placed in a series of tanks swung from the deck of the steamer, and so arranged that a constant flow of water from a cheaply improvised reservoir should keep the fish in a healthy condition. However, the use of this device proved the inexperience of the agent, for, although the fish managed to thrive for about twenty days' time, one after another died until, on the twenty-eighth day of the voyage, on landing in San Francisco, he was obliged to wire me that not a single fish had survived the passage. My answer was: "Take the same steamer back