March 1, 18S6.] 



KNOWLEDGE 



149 



across a ring coutaiuing a film, and the film is broken 

 on one side of it, the thread will be suddenly snatched 

 across the ring and be dr.iwn np tightl}' against the 



Fig. 5. 



opposite side. To facilitate niani] illation, the ends of 

 the th.read may be fastened to the ends of a wire, or 



tit, (J 



thin slip of wood. On drawing out the thread it will 

 dr.iw with it a curtain of film, and will assume the 



Fig. 7. 



curve of the arc of a circle (Fig. 7). In this way the 

 rin;:; ma}- be again filled with film and the thread be 

 entirely removed. 



A bubble ma\- be blown, a rooistened rinn- applied to 

 it, and the pipe pulled away, leavitig the bubble adhering 



to the ring. The pipe may be again dipped, passed 

 through the upper part of the bubble into its interior, 

 and a second bubble may be blown thus in the interior 

 of the first (Fig. 8). 



Fig. 8. 



B}- catching a bubble on a ring, as described above, 

 and touching it with a second ring, previously moistened, 

 it will adhere to both, so that it can be drawn out into 



Fitf. 9. 



the most elegant shapes (Fig. 9), reminding us of the 

 iridescent glass vases so popular a few years ago. 



Again attaching a bubble to a ring, the air in it can 

 be drawn out bj- inverting the moath of the pipe until, 



Fig. 10. 



on pullinsr awav the jiipe, a lenticular bubble will remain 

 (Fig. 10). 



The well-known diffusion experiment with a porous 

 jar can be very nicely shown with a film. The mouth of 



