938 



SCIENCE. 



[N. S. V<iL. XV. No. 389. 



with a lamp adapted for the voltage used, 

 and, though the lamps burn out more 

 quickly, still they last for a considerable 

 time. Of the eighteen bulbs in use in our 

 laboratory we have to renew about four a 

 year, say twenty-five per cent, annually. 

 We make use of the lamps only during the 

 latter part of the afternoon in winter when 

 the days are short. Little use is made of 

 them at night. A forty-five-volt lamp on 

 a fifty- volt current wears much longer than 

 a forty- volt lamp, and gives a light much 

 less yellow than that from a lamp adapted 

 for the voltage used. For ordinary use 

 such an arrangement is satisfactory. 



jninated, and the two ground-glass surfaces 

 through which the light passes gives it a 

 very soft efiiect. 



The light thus obtained is not perfectly 

 white, but it is white enough to prove satis- 

 factory in all the use we have given it, and 

 it is very brilliant. We frequently use it in 

 preference to daylight in the demonstration 

 of minute structures, for example in the 

 study of mitosis. 



The essential features of this plan of il- 

 lumination are the diffusion of the light as 

 explained and having bulbs adapted for a 

 voltage from five to ten volts less than that 

 of the current in use. 



An evenly illuminated surface of con- 

 siderable extent is obtained in the follow- 

 ing way: First a ground glass bulb is used 

 which softens the light ; then this is mount- 

 ed in an ordinary reading globe with mir- 

 ror back and ground glass front (cf. fig- 

 ure). The mirror-backed globes are much 

 preferable to those with painted backs. 

 The soft light from the ground-glass bulb 

 is so reflected from the mirror at the back 

 of the globe that the whole ground-glass 

 front of the globe is nearly uniformly illu- 



These lamps may be mounted in many 

 dift'erent ways. We use horizontal station- 

 ary lamps between each two desks around 

 the outside of our laboratory; and in the 

 middle of each of the central tables which 

 are used by four students apiece, we have 

 a stationary bracket in which the lamj) 

 may be raised or lowered, the lamp fasten- 

 ing by a thumb screw. Professor Drew, 

 of the University of Maine, tells me that 

 he has adopted the same style of lamp in 

 his laboratory, but that he has them 



