600 F. W. GAMBLE AND F. W. KEEBLE. 



We have been experimenting recently with a more simple 

 colour-screen consisting of a deep narrow "cell/' made by- 

 clamping two pieces of plate glass vertically against a stout 

 piece of india-rubber tubing curved in the form of a U. The 

 " cell " so formed is quite water-tight. The coloured solu- 

 tions (potassium chromate, copper sulphate, and certain 

 aniline dyes) recommended by Nagel (1898) were then tested 

 severally in different strengths and in mixtures by a Brown- 

 ing hand spectroscope. Trial alone gives the strengths 

 requisite for cutting off any portion of the spectrum which 

 it is desired to eliminate. By this means either a broad or a 

 narrow strip of the spectrum can be employed. 



In many ways this form of colour-screen is more con- 

 venient than Landoit's, but it has the drawback that for 

 monochromatic work the light intensity is apt to be low — 

 lower, in fact, than that transmitted by the Landolt screen. 



For our '^ air-circulators " we now use ordinary glass 

 flasks, which have the advantage of admitting of ready mani- 

 pulation, though their curved sides make the observation 

 of the enclosed animals somewhat more difficult than it is 

 in the case of the " water-circulator " dishes, with flat plates 

 above and below. 



By means of the " circulating " apparatus we have been 

 able to keep isolated animals under constant conditions for 

 upwards of three weeks, and it is our intention to endeavour 

 to ascertain whether by its means broods of young Hip- 

 polyte from a parent of a known colour may not be reared 

 under known conditions. 



Algee, especially of the more delicate kinds, are greedily 

 eaten by Hippolyte varians. In default of these the 

 prawns will live upon dead animals and plants, and may even 

 kill and eat specimens of their own kind. 



We have made a fairly extensive trial of preservatives 

 with a view to keeping the colour of these prawns. Alcohol, 

 glycerine, formaldehyde, sugar and water, potassium acetate, 

 chrome alum, carbon tetrachloride, have been tested for this 

 purpose. On the whole, formaldehyde (formalin 2 per 



