Dugesiella hentzi (Girard). 361 



to immerse its cephalothorax. It will never fall to avail itself of 

 the water and will spend sonie ten minutes drinking. 



The best way to keep tliem in captivity is in a Square glass 

 box with a lid of wire-netting, about two inclies of earth and a dishcon- 

 taining water. This disli should always be kept clean and filled and 

 the earth should be kept moist. For food I use grasshoppers, crickets, 

 black cockroaches and large ground spiders {Lijcosidae). They eat 

 comparatively little, not more than one grasshopper in two or three 

 days in summer and scarcely anything at all in winter. For the 

 experiments in raating I used a rectangular box of clear glass, 

 having the bottom covered with white cardboard and the top entirely 

 open to enable me to handle them with perfect freedom. Great 

 difficulties were encountered in photographiug them. In order to 

 obtain clear pictures it was necessary to place the tarantulas against 

 a white background which, to avoid reflections, had to be at a 

 distance of at least a foot behind the glass box, The experiments 

 cannot be made out of doors since the slightest breeze disturbs 

 the animals. Direct sunlight has the same effect, so that I was 

 forced to photograph them in a room near a north window. White 

 screens were used to throw as much diffused light as possible on 

 the box and to cut out reflections in the glass, The reflection of 

 the camera in the front wall of the box was entirely obliterated by 

 means of a deadblack cardboard placed immediately in front of the 

 lens and having a hole of the same dianieter as the lens. The 

 LuMiERE Sigma plate was the only plate sufficiently rapid to give 

 good results. Even then I had to use a stop F. 8, which did not 

 permit of a sufflcient depth of focus. To help the matter the box 

 was made about 40 cm long and only 18 cm wide and the lens 

 was focused on the center of the box, To change the focus during 

 the experiment is quite impossible even if one has an assistant. 

 The results could be better if one used a large lens with the camera 

 placed at a greater distance, but I did not have such a lens as the 

 size of the laboratory would have precluded its use, anywaj', Since 

 an exposure could not be made longer than ^l^^, of a second, the 

 negatives were of course all under exposed and had to be intensified. 

 Not including some 6 dozen of spoiled plates, 8 successful series 

 were made which amounted in all to about seventy negatives. The 

 reproductions shown on the accompanying plates are enlargements 

 from the original negatives. 



It seems to me advisable to give a detailed description of the 



