This statement, besides throwing an interesting light on the conditions of university work sixty years ago, gives an illustration of the wide interpretation in Scotland of the term Arts. Here it has meant, since the Chair of Natural Philosophy was founded in 1577, and held by one of the Regents of the University, _Artes Liberales_ in the widest sense, that is, the study of _Litteræ Humaniores_ (including mental and moral philosophy) and physical and mathematical science. These were all deemed necessary for a liberal education at that time: in the scientific age in which we live it is more imperative than ever that neither should be excluded from the Arts curriculum of our Universities. The common distinction between Arts and Science is a false one, and the product of a narrow idea which is alien to the traditions of our northern Universities. It is to be noted, however, that the laboratory thus founded was essentially a research laboratory; it was not designed for the systematic instruction of students in methods of experimenting. Laboratories for this purpose came later, and as a natural consequence. But for the best students, ill prepared as, no doubt, some of them were for the work of research, the experience gained in such a laboratory was very valuable. They learned--and, indeed, had to learn--in an incidental manner how to determine physical constants, such as specific gravities, thermal capacities, electric resistances, and so forth. For, apart from the _Relations des Expériences_ of Regnault, and the magnetic and electric work of Gauss and Weber, there was no systematised body of information available for the guidance of students. Good students could branch out from the main line of inquiry, so as to acquire skill in subsidiary determinations of this kind; to the more easily daunted student such difficulties proved formidable, and often absolutely deterrent. It is not easy for a physicist of the present day to realise the state of knowledge of the time, and so he often fails to recognise the full importance of Thomson's work. The want of precise knowledge of physical constants was to a considerable extent a consequence of the want of exact definitions of quantities to be determined, and in a much greater degree of the lack of any system of units of measurement. The study of phenomena was in the main merely qualitative; where an attempt had been made to obtain quantitative determinations, the units employed were arbitrary and dependent on apparatus in the possession of the experimenter, and therefore unavailable to others. In the department of heat, as has been said, a great beginning had been made by Regnault, in whose hands the exact determination of physical constants had become a fine art.