ANALYSIS IN COMMUNICATION RESEARCH 493 



the production of "sclileiren" or refraction liiu-s when two Huids of 

 differing optical density are added together without mixing. One 

 liquid is contained in a flat optical cell of a few tenths cc. capacity. 

 The other is slowly introduced below the surface of the first liquid 

 from a capillary orifice. The refraction effects set up are observed 

 through a horizontally mounted microscope under controlled illumi- 

 nation. When one of the liquids is known, such observations are the 

 basis for both specific gravity and refractive index determinations. 

 So sensitive is the method that a difference of 0.0001 in the refractive 

 indices of the two liquids is still detectable. It therefore affords an 

 excellent test for purity. If, for instance, two fractions of a distilled 

 liquid are added together with the production of "schleiren," it may 

 be assumed that the original liquid was not a pure substance. 



Glass capillary tubes have shown great versatility in microwork. 

 When the quantity of material operated upon is exceptionally small, 

 as, let us say, in the case of a foreign deposit on relay contact points, 

 their use affords distinct advantages. Almost every operation can be 

 executed through appropriate adaptations of capillary technique. 

 Thus reactions may be carried out under pressure in sealed capillaries 

 when the volume of liquid is only a few thousandths cc. and the whole 

 process watched under the microscope. Distillation and sublimation 

 are processes to which capillaries are especially suited. Suspensions 

 may be centrifuged or filtered in capillaries. In the latter case capil- 

 lary attraction is the force that draws the liquid through the filtering 

 medium. 



The possibilities of capillaries may be illustrated by a practical 

 example. Tiny discolorations were found on the surface of a polished 

 silver sheet used in photocell manufacture. Mercury contamination 

 was suspected and an analytical confirmation was desired. The pro- 

 cedure was as follows: A 1-mm. capillary tube was drawn out to form 

 a pipette having a very fine tip. With this a very small drop of nitric 

 acid was transferred to one of the discolored areas under the micro- 

 scope, and allowed to act for a few seconds, after which it was removed, 

 transferred to a capsule and the excess acid evaporated. The residue 

 was re-dissolved in a small drop of water and drawn up into a second 

 capillary tube containing a few mm. of No. 32 copper w'ire. Both 

 ends were sealed and the tube heated in boiling water for a few^ 

 minutes, after which one end was opened and the liquid withdrawn by 

 means of a finer capillary. The open end was then drawn out to a 

 very fine tube of microscopic bore. The closed end was heated by a 

 microflame, gently at first to drive out moisture, then strongly until 

 the glass had completely fused about the copper wire. Heating in 



