STUDIES ON MUSEUMS AND KINDRED INSTITUTIONS. 



389 



figuvo of a prismatic ribbed ^lass plate and a single prism. The latter 

 may be had up to a consideni})le size. The flat form (multi-prism 

 plate) costs $30 to $40 a sc^uai-e yard; and the ribbed glass plates in 

 their copper frames, $20 a s(juare 3^ard. I have made a trial of a 

 window with about 5 scpiare yards of plate g-hiss in the ethnographic 

 department of the Dresden Museum. 



because I consider the subject important for museums so long as they have not 

 learned to build them with ail rooms eciually well lij^hted: 



Everybody has perhaps noticed that rooms on the ground floor, under most cir- 

 cumstances poorly lighted, in winter, when the window panes are thickly covered 

 with frost, have an agreeable, uniform light. They have perhaps also noticed that 

 at other times, when the panes are clear and dry, only a small space near the window 

 is really well lighted and the other parts of the room remain more or less in shadow. 

 In the case of the frost a peculiar diffusion of the light is noticeable, the sharp con- 

 trast between the part near the window and the imier i)art has disappeared, and 

 only very gradiaally does the intensity of the light dinnnish toward the back part of 

 the room. 



For lighting a room !)y daylight by means of its windows, the bit of open sky 

 which looks in through the windows is more important than anything else. The 



Fio.30. Fig. 31. 



amount of light derived by reflection from other objects is, on the other hand, very 

 small. The nearer opposite buildings are to the windows of a room or the higher 

 those buildings are, the smaller will be the bit of sky which can look into tlie room 

 and light it up. Why, then, under the same conditions of light, does such a room 

 ai^pear ))righter — that is, more uniformly lighted — when the windows are frosty than 

 when they are clear? 



The window is struck by a bunch of rays which goes through the plane i)arallel 

 glass with(nit changing their direction (fig. 30), and the more obliquely they fall 

 the smaller will be the surface of floor which they strike and illuminate. The light 

 which illuminates the rest of the room is derived by reflection from this small spot. 



The particles of frost on tiie windows consi.st of crystallized water, of an iuunense 

 nui^iber of small hexagonal prisms. Every ray of light which strikes one of these 

 small prisms is refracted and takes another direction. Since the nund)erless i)risms 

 lie in every conceivable direction on the surface of the glass, the rays of light strik- 

 ing them are also turned in every possible direction. They will not fall parallel (ui 

 the floor as l)efore, but shoot into the room in every direction. The room is no 

 longer filled with the scanty reflected light which comes from the lighted sjjot on the 

 floor, but with light direct from the sky (fig. 31). 



This observation has led to a discovery which is of the greatest importance in 

 lighting dark I'ooms liy daylight. 'I'hc pi'oblcm to solve was nu'rcly to replace the 



