]\gt ; 464 Mr. W. H. Preece .
On the [ Apr. to ontbalance all the weighty chemical and physioel difference between them The nature of the evidence derived bom the spectroscope is hero quite misrepresented ; probably it was not well understood because its novelty .
I have fonnd that the spectra of magnesium , zinc , and cadmium are the result of three series of harmonic vibrations with similar intervals , the fundamental vibrations of which differ only in pitch .
I have made some similar observations with regard to copper , silver , and mercury , aluminium , indium , and thalfium , calcium , strontium , and barium , though maps of some of these spectra have not yet been published .
I believe that in series of elements snch as { bese , which exhibit tional differences in properties and in the properties of their compounds , and approximately equal numerical differences in their atomic weights , we are dealing with the same kind of matter in different states of condensation , or in ot , her words , with matter having similarly constituted molecules , the vibrations of which are in the same direction and at similar intervals but with diiferent velocities .
If we attempt to classify beryllium in accordance with the views of Nilson and Pettersson , the elements scaudium and yttrium , with atomic weights 44 and 89 respectively , must yield spectra characteristic of the series of which aluminium is the first member , but it is not possible to find a place for beryllium in this group , nor in those to which cerium , lanthanum , and didymium belong ; it is , in fact , by a process of exclusion first and selection afterwards that the element falls into the dyad series .
In the position which I assign to beryllium , we can account for it being related , through the properties of its compounds , to magnesium and zinc on the one hand , and to aluminium on tlJe other .
If there is one consideration of greater importance than another which should lead to the determination of the position of an element in a series , it is the mode of vibration of the molecule , and of that we have evidence in the spectrum .
When the element is one with a low atomic weight there is no difficnlty in interpreting such evidence .
Il .
On the Heating Effects of Electric Currents By WILLIAM PREECE , F.R.S. Received March 19 , 1884 .
The production of heat in electrical conductors due to the transference of electricity through them has become a matter of very great practical importance .
A knowledge of the variation of the law , due to the dimensions and character of the conductor , is essential for lightning protectors and for the leads of electric lights .
1884 .
Heating Effects of Electric Currents .
Atmospheric electricity has proved a great danger to insulated wires , subterranean and submarine , and to telegraphic apparatus generally .
Not only do the direct discharges of atmospheric electricity enter the wires , but very powerful currents are induced in neighbouring wires when { hese discharges take place , either between cloud and cloud or between cloud and earth .
Various plans have been devised to protect apparatus and wires from these currents .
Lightning protectors based on effect of points , on the facility of discharge through vacua , on the low resistance of thin air-spaces to high potential , and on the fusibility of thin wires , have been used .
The most careful and long-eoffiinued observations have shown that the survival of the fittest is found , in the use of a thin flat air-space ( fig. 1 ) , supplemented by a fine well insulated protecting wire , , of high resistance , wound around a brass rod in connexion with earth .
The pace is obtained by superposing two smooth plane surfaces of hick brass , separated each other by a space of 004 of an inch , by means of a frame of mica or paraffined paper .
It requires ( according to the researches of Messrs. Warren De La Rue and Hugo confirmed by my own observations made in .
Warren De La Rue 's laboratory , a potential of 800 olts to strike across such an air-space , and when this is done a path of no resistance , according to Faraday , is established for the atmospheric discharge to flow to earth .
But destructive currents of induction are often produced in telegraph wires , which have not the requisite potential to strike across this air-space , and the element of time may enter to force some of the current into the cable or apparatus so as to * ' ' Phil. Trans Part II , vol. 174 .
Dr. De La Rue has been kind enough to repeat these experiments the foUowing resuIts : \mdash ; Air-space .
Potential .
001 in . . . . . . . . . . . . . . . . . . . . . . . . . . . .
453 volts .
, . . . . . . . . . . . . . . . . . . . .
, . . . . . . . .
659 \ldquo ; . . . . . . . . . . . . . . . . . . . . . . . .
005 H. P. , , 1884 .
466 Mr. W. H. Preece .
On the [ Apr. it .
Hence a wire bas been added , which , while it will allow all working currents to pass without retardation , will not allow a current strong enough to damage the cable to flow through without its being fused .
It thus acts as a safety valve .
The length and dimensions of this wire have not been hitherto determined by any stringent conditions .
After having ermlned he oonditions required to give the maximum effect with the -caIled ' ' plate pro .
tector beca1ne anxlou find out the conditions that would a similar result with the " " wire prot ector.\ldquo ; Now , the very strongest current that can , under any possiblo circumstances , enter a wire from the working batteries is one of 500 milliamperes nnit current , while atmospheric currents enter the wires of all strengths from 1 milliampere to 30 or 40amperes , or even more .
Therefore , it was desirable*o find that wire which woutd not fuse with 500 milliamperes , but would fuse with currents of about 700 milliamperes and over .
Hence , I took a source of electricity , which was a large of Clamond 's make , which would give the necessary current , rbeostat by hich it was possible to regulate this onrrent to any strength , a Thomson mirror galvanometer which ) bled me to measure , record , and reguJate every change of current , and a discharger by which I was abl hold , measure , and adjust the wirea to be tested .
1 .
With a given current and a uniform wire of given diameter and given material , the effect , allowing for the cooling effect of the binding screws , was independent of the length , tor whether the length experimented upon was an inch , a foot , or a yard , it always fused when the current reached a certain definite strength .
The point of fusion was irregular .
for it simply depended on some irregularity in the uniformity of the wire , and weakest link in the chain went .
2 .
Samples of platinum wire of various diameters , each of 6 inches in length , were taken , and the current increased gradually until the wire fused .
The results are given in the following table:\mdash ; * British , 1880 .
Heating ffects of Electric Currents .
467 The above figures are the average of numerous experiments , and they are plotted out in fig. 2 .
The general law governing the ratio between the current strength 1 ' and the diameter of the wire when the latter is raised to a definite temperature , and where radiation is free and uncheckod by any ather1 ' manous envelope , appears to be that the current should vary as the ' diameter diameter ; this may be proved as follows .
The heat developed in any cylindrical wire is by Joule 's law\mdash ; where being the specific resistance , the length , and the diameter , then\mdash ; ; or if we maintain the ratio between the current and the diameter constant , the same amount of energy will be wasted in the conductor .
But the temperature of the conductor is dependent on the rate at which the heat is radiated and conducted away , and when the heat imparted by the current just that lost by radiation * ' The Electrical Review June 24 , 1882 , p. 454 , and Prof. George Brit. Assoc. , 1882 .
VOL. XXXVI .
2 468 Mr. W. H. Preece .
On the [ Apr. and convection , the temperature becomes constant .
But the heat lost val'ies with the surface , and the smaller the surface the less loss , and therefore the temperature will vary with the diameter .
Hence as\mdash ; and also therefore , or , The results of the foregoing experiments on the current to fuse wires of different diameters would appear to contradict thit law , and to show that the currenl varies as the diameter , but are several elements in the method of making the measurements which tend to modify the general law .
Platinum especially those of small diameter\mdash ; are liable to flaws which practically reduce their effective diameter to a great extent ; also the larger wires from their greater weight necessarily tend to part a@under at a lower peratnre than those which are lighter and on which the strain is less .
3 .
In thermometric measurements there are only two fiducial points usually employed , , the freezing point , and the boiling point , but there are two other points equally well marked , though perhaps not quite so accurately fixed .
These are the points of selfluminosity and of fusion .
The first point has been determined by Daniell to be , and by Draper at , and it has been shown by the latter observer that this is true for all substances .
I was anxious to discover whether the law before stated would apply in the case of the production of the lowest visible red radiations by currents .
I therefore took several wires of different materials , and of different diameters , but all of the same Iength , and by observing them in a carefully constructed dark chamber , determined the currents that produced self-Inminosity , these currents being measured by finding the difference of potential at the ends of a thick German silver wire ( R ) whose resistance was , inserted in the same circuit .
The results of the experiments , which were very carefully made , are given the following tables:\mdash ; 1884 .
Effects of Electric Currents .
Constant defiection from standard Daniell cell through 10,000 ohms . .
Constant defiection from standard Daniell cell On the Heating Efects of Electric Currents .
[ Apr. Constant deflection from standard Daniell The resuIts are plotted out in fig. S. It will be seen that the observed and calculated results agree to a considerable degree of accuracy except in the case of platinum , which behaves as in the previous experiments , and is generally found to be egular in its quality .
rhese experiments were made upon wires exposed to the air , where radiation is free .
I am anxious to repeat them upon wires covered in insulating material and buried in the ground , but I have not been able to do so up to the present mom .
The law with reference to such wires has a very important bearing on the size of electric light 384 .
] Spectroscopic Studies on Gaseous Explosions .
leads , for it shows the necessity of making them large enough to prevent the possibility of their being heated above normal temperatures , otherwise points of danger are very easily reached by increments of current .
III .
: ' Spectroscopic Studies on Gaseous Explosions .
No. I By G. D. LIVING , M.A. , F.B.S. , and JAMES DEWAR , M.A. , F.R.S. , Professors in the University of Cambridge .
Received March 28 , 1884 .
Having occasion to observe the spectrum of the flash of a mixture of hydrogen and oxygen fired in a Cavendish eudiometer , we were struck by the brightness , not only of the ubiquitous yellow sodium line , but of the blue calcium line and the orange and green bands of lime , as well as of other lines which were not identified .
The eudiometer being at first clean and dry , the calcium must be derived either from the glass or from some spray of the water over which the gases with which the eudiometer was filled had been confined .
It seemed incredible that the momentary flash should detach and light up lime from the glass , but observations have pointed to that con-

