Table 1. --Stains tested 



permitted easy injection without excessive drip 

 and allowed for ready control of penetration 

 depth . 



supplied for the duration of the experiment. 



Controls 



G . Feeding : Mullet, cut as for normal 

 feeding, was placed in a stain solution for 15 to 

 20 minutes and agitated to ensure even staining. 

 The solution was then decanted, and the food 

 was washed in sea water to remove excess 

 stain . The stained food was then dropped into 

 the tank following normal feeding practice. After 

 a number of stained feedings, normal food was 



The prime requisite of a shrimp stain is 

 that it must provide a fast and easily distinguished 

 color. Most of the experiments were designed 

 to test staining characteristics, and regular con- 

 trols were not established. Simultaneous controls 

 were run with the more successful stains. In 

 control experiments an equal .lumber of untreated 

 shrimp were held in a separate compartment in the 



