260 ON THE ACTION OF LIGHT 



Supposing then that the situations in which the two 

 patches of sponge were respectively placed, ( one being ex- 

 posed to the light, and the other deprived of it), formed the 

 real cause of the difference of colour in them, I took home the 

 tile in a vessel of water, and commenced the following 

 experiments, in order that I might, if possible, prove the 

 coiTectness or falsity of my conjecture. 



On the 1 8th August, 1837, the day in which I obtained 

 the fragment of the tile, I lost no time in returning home 

 with it, carrying it in a vessel filled with water, very gently, 

 and with great caution, lest the living specimens of the 

 sponge should be in any way shaken, or broken, or otherwise 

 injured. I then placed the tile in a basin of fresh water, but 

 exactly reversed the position of the two patches of sponge, 

 that is to say, I put the underside of the tile with the pale 

 hrown piece of sponge growing to it uppermost, and exposed 

 it to the light in a window, where I allowed the whole rays 

 of the sun to enter and shine upon it ; and the original upper- 

 side of the tile, containing the green patch of sponge, of 

 course, then occupied the underside, and was almost entirely 

 deprived of light. I changed the water twice or thrice a 

 day, adding it fresh from the pump, and not rain water, or wa- 

 ter taken from a pond, or in any way intermixed (as far as I 

 could possibly ascertain), with any fresh vegetable or colour- 

 ing matter. After a couple of days, I thought the brown 

 piece of sponge began to assume a greenish tint, and the 

 green piece under the stone to lose, though in a less degree, 

 somewhat of its deep colour ; these appearances, after a few 

 days more, became distinctly manifest, and by the 29th 

 of August, the brown or sand-coloured patch of sponge had 

 changed to a bright grass-green, and the green patch on the 

 under surface of the tile had diminished in its gi'een 

 hue, and had approached to a hght grey. At length, 

 after the expiration of twenty days, the upper sponge had 

 much increased in depth of colour, and the lower one had 

 lost a good deal of its green, (2) and had changed to a light 

 brownish green, resembling, indeed, the colour of the patch 

 when originally growing to the underside of the tile, as I 

 had first observed it upon taking it out of the rivulet. Hence, 

 I conceived it to be evident, that the action of light directly 

 caused the green colour to be secreted in this sponge ; and 

 for the sake of corroborating this opinion, I subsequently 

 made these additional experiments. 



A short time afterwards, I fished up by means of a wire- 

 gauze net fixed to the end of a rod, a large mass of the river 

 sponge fi'om the same rivulet ; this specimen was growing 



