Prof. H. L. Callendar on Platinum Thermometry. 221 



tion, which will be described in a subsequent communication. 

 The general result of any residual thermal effects which may 

 be present is to produce a change in the apparent value of d, 

 since the thermo-E.M.F. follows approximately a parabolic 

 formula. It is possible, for this reason, to obtain consistent 

 and accurate measurements of temperature with a platinum 

 thermometer in spite of large thermal effects, but the value of 

 d would be very considerably affected. 



On the " Vanishing Temperature" — There appears to be a 

 very general consensus of opinion, based chiefly on the par- 

 ticular series of experiments which are under discussion, that 

 the resistance of all pure metals ought to vanish, and does 

 tend to vanish at a temperature which is no other than the 

 absolute zero. If, hoAvever, there is any virtue in the para- 

 bolic method of reduction, it is quite obvious, on reference to 

 the column headed " Vanishing Temperature " in the above 

 table, that the resistance "tends to vanish" in the case of 

 most of the common metals at a much higher temperature. 

 The vanishing temperature f is the value of t deduced from 

 the fundamental zero pt° in each case by means of the 

 difference-formula, employing the value of d° given in the 

 table. The most remarkable metals in this respect are pure 

 copper and iron, which tend to become perfect conductors at 

 a temperature of —223° approximately, a point which is now 

 well within the experimental range. These are followed at a 

 very short interval by aluminium, nickel, and magnesium. 

 In the case of copper and iron special experiments were made 

 at a temperature as low as —206° 0., at which point the 

 rate of decrease of resistance showed little, if any, sign of 

 diminution. The exact value of the vanishing temperature 

 in each case is necessarily somewhat uncertain owing to the 

 necessity of extrapolation, and also on account of possible 

 uncertainties in the data ; but there can be no doubt that the 

 conclusion derived from the formula represents, at least ap- 

 proximately, a genuine physical fact. Whether or no the 

 resistance does actually vanish at some such temperature may 

 well be open to doubt. It would require very accurate ob- 

 servations to determine such a point satisfactorily, as the ex- 

 perimental difficulties are considerable in measuring so small 

 a resistance under such conditions. It is more probable that 

 there is a singular point on the curve, similar to that occur- 

 ring in the case of iron at the critical temperature, at which 

 it ceases to be magnetic. It is also likely that the change 

 would not be sudden, but gradual, and that indications of the 

 approaching singularity would be obtained a few degrees 

 above the point in question. Below this point it is even pos- 



