Phyllis Harsh returns a stranded "baby Caddy" to the sea, Johns Island, 1991

Off the coasts of British Columbia and Washington, sailors and fishermen have long told of a sea serpent. Its scientific champions call it Cadborosaurus. Most people call it Caddy. Johns Island lies in the San Juan Islands of San Juan County, Washington, just east of Stuart Island and north of Spieden Island. Five residents in the 2000 census, a children's camp at the west end, private cabins at the east.

July 1991. No exact day survives. Phyllis Harsh, a resident of the island, was near her home. On the beach lay a small animal, about two feet long, roughly 60 centimetres. Harsh called it a "baby dinosaur."

It was stranded. Harsh took a tree branch. With it, she lifted the animal back into the sea. It "was able to make its way back into deeper water." Nothing more is recorded about its colour, its eyes or any sound it made.

Harsh was the only witness, and this was not her only story. She also stated for the record that she had seen a full-size Caddy in John's Island Passage in 1990. And she said she had found what appeared to be a "dinosaur" skeleton beneath an eagle's nest, also on Johns Island.

No newspaper covered it in 1991. No photograph was published. No specimen was kept. Every trace leads back to Harsh's own account, written up by Bousfield and LeBlond in Cadborosaurus: Survivor from the Deep, published in 1995 and again in 2000. The account may first appear in their formal description of Cadborosaurus willsi in the journal Amphipacifica, though that remains unconfirmed.

The retellings drifted. Some say she found the animal beached. The encyclopedia version says she "caught" a baby Caddy. A 2025 web article went further, adding a net, a blurry photo, a furry neck and flippers, none of them in any earlier source. Joseph Reigel's 2025 book even moves the scene to "Little Johns Island."

Sceptics point to the pipefish. The idea grew from the only other claimed live juvenile, caught by Hagelund off De Courcy Island in August 1968. In 2011 Woodley, Naish and McCormick identified that animal as a bay pipefish in the Journal of Scientific Exploration. Bousfield and LeBlond answered with "Pipefish or pipe dream?"

A 2012 note on Frontiers of Zoology suggested both juvenile incidents involved small fish such as pipefish. The peer-reviewed paper itself, though, deals only with Hagelund. No one has published another candidate for the animal Harsh saw.

The Salish Current put the problem plainly in 2025. "Other than Harsh's account, there is no evidence of this."

The case lives on in Michael Newton's Hidden Animals and Encyclopedia of Cryptozoology, in Reigel's Mysterious San Juan Island and in the Cadborosaurus pages of the web. It holds a rare place in sea serpent lore. Not a distant hump on open water. A small animal on a quiet beach, lifted back into the sea by one islander with a branch.