EXPERIMENTS ON EXPANSION. 759 



ef, should be filed very smooth, and slightly but uniformly convex, 

 as indicated in the figure. 



The heating of the vessel with the paraffine oil must be done very 

 cautiously, as the glass is apt to crack, and very serious accidents 

 might be caused if the paraffine, which is very inflammable, were to 

 take fire. The safest way is not to heat the vessel by means of a 

 lamp, but by immersing it in a pot containing hot water. The 

 expansion is still better observable if a small flask with a narrow 

 neck is filled with the liquid, so that it only just fills the wide 

 portion of the flask up to the neck, and the flask is heated. In this 

 experiment, however, it is better to use spirits of wine instead of 

 paraffine oil, because the latter cannot well be entirely removed from ' 

 the interior of a vessel without rubbing off the adhering drops with 

 a piece of cloth or blotting-paper, which is impossible with a 

 small flask. The spirits of wine should also be heated with great 

 caution, and not too strongly. Water expands less than either 

 paraffine oil or spirits of wine. If the experiment is to be made with 

 water, the flask, about 6 or 8 cm wide, should be closed by a well- 

 fitting cork, perforated for receiving a glass tube, 3 mm wide, and 

 20 or 30 cm long. Common water always contains a quantity of air 

 in solution, and when heated the air escapes in bubbles ; the water 

 used for the experiment should therefore be first boiled, and then 

 allowed to become cool again. The flask is filled up to the neck, 

 so that no air may remain in the neck when the mouth is closed 

 by the cork ; if the water rises too high in the tube, some of it may 

 be removed by sucking through a straw, which is introduced into 

 the tube, until the water, stands a few centimetres above the cork. 

 The heating in this experiment also is best done by immersing the 

 flask in hot water. At first, as soon as heat is applied, the water 

 seems to contract, for it falls a little in the tube ; but the cause of 

 this is that the sides of the vessel are heated first, before the heat 

 reaches the water. The vessel therefore becomes of somewhat 

 greater capacity than it was previously, while the water really pre- 

 serves its original volume, and cannot reach up to the previous 

 height in the larger vessel ; hence it falls at first. As soon as the 

 heat reaches the water itself it begins to expand, and rises in the 

 tube. 



The only precautions necessary in the experiment on the expansion 

 of air are to avoid the heating of the neck of the vessel, otherwise 

 it is apt to crack when the water rises in the neck after the flame 

 is withdrawn ; and further, not to hold the flame in one place under 

 the vessel, but to carry it constantly round and round, so that the 

 heat may be applied uniformly and slowly. 



