Pelagic fishes - A Fascination for Fish: Adventures of an Underwater Pioneer - page 280

to have excellent vision in the blue range of the spectrum, and we hoped
that they would easily see—and avoid—the pattern of the blue tiles.
An absolutely evenly colored blue wall with no visual irregularities could
be very dangerous to these fast-swimming fish. It was critical for them
to know just where the wall was.
A number of years ago, at the o¤shore island of Socorro in Mexico,
I had taken underwater photographs that captured the beautiful blue
color of open ocean water. We sent these photographs to a tile manu-
facturer, Mosaicos Venecianos de México in Cuernavaca, to see how
closely they could match the di¤erent shades of blue. The samples they
sent back were very close to the colors of the open ocean. A test panel
was made up with nine shades, from light blue at the top to deep blue
at the bottom. We hung this panel in the Monterey Bay Habitats ex-
hibit and looked at it through thirty feet of water. It looked great, and
the breaks where one shade of blue changed to the next weren't visible
at all. The tiles, we were confident, would work just fine in the Outer
Bay exhibit, whose curved back wall would be forty to fifty-five feet
away from viewers. We decided to place our order: 1,336,000 tiles, each
one three-quarters of an inch square.
The tiles were delivered, and the Kreysler Company sta¤ began the
installation. Because the smooth, sloped liner inside the tank made it
impractical to set up sca¤olding, the tile was applied from two float-
ing rafts. The installers worked from one end of the tank to the other,
laying tile in three-foot-wide bands. As each level was finished, more
water was added to the tank to float the rafts a little higher.
About three-quarters of the tile had been applied when we decided
to fill the tank all the way and take a look at it. To our horror, we saw
that even through fifty feet of water the di¤erent bands of blue were
quite visible. What had looked great on our small-scale test panel did
not work at full scale.
This was an awful development. We tried staggering the tiles where
one color met the next to see if the bands would disappear, but with
no luck: they were almost as obvious. The banded appearance of the
back wall was simply unacceptable. We decided that a single, uniform
blue was our only option. It was certainly better than horizontal stripes.
This correction was going to be an expensive hit on the budget, and