TRANSACTIONS OF THE SECTIONS. 41 



metal, or of crystals arranj^ed axially and eq^uatorially, the current being then sup- 

 ported by the i?'eltier absorption and evolution of heat. Now discontinuity is the 

 greatest possible change which can occiu- in the molecular structui-e, and it there- 

 fore appears not improbable that the cuiTents due to loose contacts, or in other 

 words to discontinuity, may be referred to the same cause. as the currents due to 

 varying temper or to crystalline structure, — that is to say, to absorjDtion of heat 

 where the change of structure occurs, heat being evolved in other parts of the cir- 

 cuit at a lower temperatui-e. I hope soon to decide this question and others by 

 further experiments in various media, with definite pressiu'es at definite tempe- 

 ratures. 



When various metals are combined, striking effects are produced, the sti-ength 

 of the common thermo-electric current from the joint being often increased fifty or 

 a hundred fold when loose contacts are substituted for tight contacts. The direc- 

 tion of the ciurent is also frequently reversed. 



The results of these combinations are necessarily complicated, and require further 

 experiment and analysis before publication. In the ordinaiy thermo-electric battery 

 made from pairs of dissimilar metals, a very small proportion of the heat commu- 

 nicated to the joint is converted into electricity, which is therefore obtained from 

 them at a great disadvantage. But considering the comparatively great intensity 

 of the currents produced when loose contacts are adopted, it seems possible that by 

 their means a considerable part of the heat used may be absorbed in the production 

 of electricity, which woidd in that case be more cheaply obtained from heat, than 

 directly from chemical action. 



It is needless to allude to the consequences which would ensue should a cheap 

 source of electricity be discovered ; but without anticipating such important con- 

 sequences from the discoveiy of the loose contact cun-ents, they certainly seem a 

 fit subject for further investigation. Meanwhile it is interesting to consider how, 

 when two wires are tightly joined, the lieat given them by the flame travels but a 

 few inches slowly along them, producing all its sensible effects on objects in the 

 immediate neighbom-hood ; whereas when those wires are moved asunder to an 

 almost imperceptible distance, that same heat may in an instant be flashed as elec- 

 tricity through thousands of miles, reappearing distributed once more in the form 

 of heat abnost sunultaneously in every part of the whole circuit. 



On the Secular Changes of Terrestrial Magnetism, and their Connexion with 

 Disturbances. By the Eev. H. Lloyd, D.D., D.C.L., F.B.S., M.B.I. A. 



Of the various changes to which the dii-ection and intensity of the earth's mag- 

 netic force are subject, unquestionably the most mysterious are those which, from 

 their analogy to the slower changes of the solar system, have been denominated 

 secular. No one has, as yet, ofl'ered even a plausible conjecture in explanation of 

 these phenomena; while, on the other hand, it has been felt by all who have 

 studied them, that their causes lie so deep, and are so closely connected with the 

 hidden nature of the force itself, that the knowledge of them would, in all proba- 

 bility, imlock most of the secrets of terrestrial magnetism. For these reasons, any 

 attempt, however imperfect, to add to our knowledge of the laws which govern 

 them will probably be received with indulgence by magneticians. 



It has long been known that, in addition to the changes which pass through their 

 whole cycle of values in a day, or in a year, and which are thence called periodic, 

 the magnetic elements at a given place are subject to changes of another kind, 

 which continue for a long time in the same du-ection. It has been generally sup- 

 posed that, for a limited number of years, the rate of these changes at any given 

 place was either uniform, or else uniformly accelerated or retarded; so that they 

 could be mathematically represented by a fonnula consisting at most of two terms, 

 one of which was proportional to the time (measured from some certain epoch), 

 and another to its square. In other words, it has been supposed that the 7nean 

 yearly values of the magnetic elements were subject to no Jluctuations oi minor 

 period. 



This view, so far at least as concerns the secular changes of the inclination, has 

 been completely disproved by Professor Hansteen. From the long and accurate 

 series of observations of this element made by himself at Christiania for more than 



