72 Royal Society : — 



covered with watch-glasses, and allowed to cool ; in one of the 

 beakers a small glass bucket, attached to a thread, had been placed 

 and boiled up with the solution. Next, both beakers were arranged 

 under a large bell-jar, and the silk thread from the bucket passed up 

 between the stopper and the neck of the jar. The solutions were 

 then uncovered, after waiting ten minutes for any nuclei which might 

 have been disturbed to fall ; a fine wire was passed down into the 

 beaker containing the bucket, and as far as possible from the part 

 of the solution through which it would pass on being drawn up. 



The bucket, now full of the crystallized normal sodic sulphate 

 (Na 2 S0 4 , 10H 2 O), was raised, and lowered into the second beaker 

 of still fluid solution ; immediately that the point of one of the crys- 

 tals hanging from the under surface of the bucket touched the solu- 

 tion, crystallization was set up instantaneously throughout the mass. 



This experiment was performed many times, and with every pos- 

 sible care to prevent the entrance of nuclei other than those pur- 

 posely borne by the wire. 



A modification of the above plan was tried and with similar 

 results. 



Exp. A tubulated glass bell was fitted with a cork bearing two 

 glass tubes, open below and closed above with cotton-wool ; they 

 were bent so as to permit both of them being placed in one and 

 the same beaker, or into either separately. 



In the first place, the ends of the tubes inside the bell were freed 

 from nuclei by passing them through a flame ; two beakers of cold 

 supersaturated solution were then placed in position under the bell- 

 jar, and their covers removed. After waiting five minutes or so for 

 any dust to settle, both tubes were next lowered into one of the 

 beakers, on opposite sides, so as to be as far apart as possible. A 

 dirty wire was now passed down one of the tubes, when, of course, 

 crystallization immediately took place, and was propagated across the 

 beaker. The second tube, with its adhering crystals, was then 

 raised and lowered into the second beaker, when, the moment the 

 extreme point of the longest crystal touched the surface of the so- 

 lution, crystallization immediately started from that point, and the 

 whole contents became solid. 



A third variation was then made in this experiment. One of 

 the two beakers was replaced by a U-tube of thin, hard glass, one 

 of the before-mentioned tubes being inserted into either limb. Crys- 

 tallization, when set up in one limb, travelled round the bend and up 

 into the other, from which crystals were transferred, as before, to a 

 beaker or flask of solution also under the bell-jar. 



The three modifications of this form of experiment were tried time 

 after time, and always with the same unvarying result. Solutions 

 which were supersaturated although not perfectly, and therefore 

 less sensitive, were operated upon in this way ; but, even with such 

 less favourable circumstances, the normal crystals always started 

 crystallization in the solution to which they were added. 



To ascertain, if possible, whether nuclei, other than crystals of the 

 normal salt, were carried by the tube or its adhering crystals, a 



