402 Profs .
G. D. Living and J. Dewar .
[ June 16 vibrated , not during , but after , the changes of load through which the given state has been reached .
It is also shown that this approximate equality produced by vibration continues after the vibration ceases .
Also , that when a cycle of loads is gone through , afterwards , without vibration , the old difference between the " on " and " off " curves reasserts itself .
It is suggested that the cyclic phenomenon so conspicuous in this investigation is not peculiar to the thermoelectric effects of stress , but is probably present in other effects of stress , and may perhaps be found to occur in the .
changes of any quality of matter which is a function of another variable quality ( such as temperature ) when the latter quality is subjected to increment and decrement .
Lastly the results of certain independent experiments made by others in other branches of physics are referred to in confirmation of this suggestion .
XII .
" On the Reversal of the Lines of Metallic Vapours .
No. VIII .
( Iron , Titanium , Chromium , and Aluminium .
) " By G. D. Liveing , M.A. , F.R.S. , Professor of Chemistry , and J. Dewar , M.A. , F.R.S. , Jacksonian Professor , University of .Cambridge .
Received June 2 , 1881 .
In our last communication on this subject we observed ( " Proc. Roy .
Soc. , " vol. 29 , p. 405 ) that iron introduced as metal , or as chloride , into the electric arc , in a lime crucible , in the way which had proved successful in the case of many other metals , gave us no reversals .
We succeeded , however , in reversing some ten of the brightest lines of iron , mostly in the blue and violet , by passing an iron wire through one of the carbons , so as to keep up a constant supply of iron in the arc .
Considering the great number of iron lines , and that so many of them are strongly represented amongst the Fraunhofer lines , it seemed somewhat surprising that it should be difficult to obtain a reversing layer of iron vapour in the arc inclosed as we use it in an intensely heated crucible .
A like remark might be made respecting titanium , which is almost as well represented as iron in the Fraunhofer lines , but has heretofore given us no reversals .
Almost the same might be said of chromium , except that the number of chromium lines is so much less than that of either of the other two metals .
We have since found that most , if not all , of the strong lines of these three metals may be reversed by proper management of the atmosphere and supply of metal in the crucible .
Indeed , with regard to iron we have found that the method employed with other metals was successful so far as the ultra-violet rays were concerned , though it 1881 .
] On the Reversal of the Lines of Metallic Vapours , 403 failed for less refrangible rays .
^Then iron has been put into the crucible through which the arc of a Siemens ' dynamo-electric machine is passing , and then fragments of magnesium dropped in from time to time , most of the strong ultra-violet lines of iron are reversed .
The magnesium seems to supply a highly reducing atmosphere , and to some extent carry with it the iron vapour .
It also produces a good deal of continuous spectrum , at least in certain regions , and against this the iron lines are often depicted on the photographic plates sharply reversed .
In this way we have observed the reversal of the strong iron lines about the solar lines L and M , four strong lines below N , the line O , all the strong lines from Sg to U inclusive , and two strong groups still more refrangible .
Potassium ferrocyanide introduced into the arc instead of magnesium gives a reversal of the .
same lines as are mentioned in the foregoing paragraph .
Iron wire fed in through a perforated pole gives reversals of the highest group ( wave-length 2492 to 248(1 ) , but with the lines so much expanded as to form broad absorption-bands instead of lines .
With a vertical arrangement of the carbons , similar to that used by us for resolving the fine double lines ( u Proc. Roy .
Soc. , vol. 32 , p. 231 ) , and a stout iron wire in the axis of the lower ( positive ) carbon , many more lines in the visible part of the spectrum are seen expanded and reversed .
This effect is sometimes enhanced by leading into the crucible through the upper carbon , which is perforated for the purpose , a very gentle stream of hydrogen gas ; the stream must be no more than is just sufficient to give a tiny flame at the mouth of the crucible ; a stronger stream diminishes the amount of metallic vapour , probably by its cooling action , and lessens the effect .
By this treatment , some of the strongest lines of iron remain reversed for some time , the weaker lines are seen to expand and be reversed for a few seconds at a time , when , from a change in the intensity of the current , or some other reason , a larger amount of metal is volatilized and shows itself by burning in brilliant scintillations at the mouth of the crucible .
A list of the iron lines reversed , designated by their approximate wave-lengths , is subjoined .
5614*5 5585*5 5445*6 5428*5 5370*5 5327 the more refrangible of this pair .
5268*5 E Profs .
Gr .
D. Living and J. Dewar .
[ June 16 5226 the less refrangible of this double line .
5166-8 4956-5 both lines of pair , the less refrangible strongly 4919-7 3727 M 3007-5 2744-7 4918 3722 3006-5 2743 4890-5 3720 3000 2741-3 4890 3707-5 2999 2736-2 4871 both lines of pair .
3705-5 2994 t 2722-7 4414-5 3647 2984 2720 4404 3631 2982 2718 4382-5 3502 2974 2549-5 4325 3492 2971 2545-5 4314-5 3475 2967-5 2541 4307 3465-5 2965-5 2535-7 4271 both lines of pair .
3440 0 2957 2534 4071 3402-5 2953-5 2529 4063 3272* 2948 U 2528-6 4045 3246* 2941 2527-3 3898 3099-5 S2 2937 2524-2 3886 3082 2932 2523 3859 3072 2929 2518-3 3856 3057 2912 2511 3833-5 3056 2871 2502 3827-5 3046-5 8 2866-3 2492 3825 3041 2860-8 2491-6 3824 3036-5 2836-4 2491 3820 L 3030 2829-8 2489-3 3815 3024-5 2824-2 2485-4 3767 3022-5 2767-3 2484-6 3763-5 3020 T 2755-2 2481 3757-5 3017-5 2754-3 2480 3745-5 3016-5 2749 2474 3736-5 3008-5 2748 2465 3734-5 It is by no means always the strongest lines which are reversed .
For instance , in the group of lines between wave-lengths 5428*5 and 5396 there are four strong lines 5423"5 , 5414'6 , 5403 , and 5382"4 , which are all much expanded and diffuse when iron is introduced into the arc , but have not been seen reversed by us , while three other lines \#151 ; 5396 , 5405 , and 5428"5\#151 ; are well reversed , but not much expanded .
* It is possible that these may have been the copper lines which were reversed , and not the iron lines , which are nearly identical in position .
1881*3 On the Reversal of the Lines of Metallic Vapours .
405 Considering the enormous complexity of the iron spectrum , which probably reaches to something like a thousand distinct vibrations , we should expect to find a greater variation in the relative strength and reversibility of the lines of this metal , according to the circumstances in which it is placed , than we find in the case of metals giving simpler spectra .
So that we may say , on the whole , that the phenomena in the case of iron are of much the same general character as we have observed before in the case of other metals ; but the number of observations to be made on so extensive a subject renders it extremely difficult to come to any general conclusions .
When the perforation of the lower carbon is filled with titanium cyanide instead of the iron wire , the titanium lines come out very brilliantly and steadily , and many of them , especially in the green and blue parts of the spectrum , are expanded and reversed .
We give a list of those we have observed to be reversed .
In the case of chromium , introduced into the crucible either as oxide or as bichromate of ammonia , there were no reversals until a gentle current of hydrogen or of coal gas was led in through the perforated carbon .
This brought out the triplet in the green , wave-lengths 5207 , 5205 , 5203 , sharply and steadily reversed , and likewise the three strong lines in the indigo , wave-lengths 4289 , 4274 , 4253 ; also a triplet near N at wave-lengths about 3578 , 3593 , 3606 , apparently coincident with strong lines in Cornu 's map of that part of the solar spectrum , and a rather strong double line just below O at about wave-length 3446 .
The reversal of another chromium line at about wave-length 3217 is doubtful .
A triplet at wave-lengths 2799*8 , 2797,2794 , is more easily reversed than any other of the chromium lines .
This triplet is generally strongly developed whenever a compound of cnromium is introduced into the crucible , so that we conclude that it is due to that metal , but it is sometimes visible in the photographs when other chromium lines are not seen .
A still more refrangible chromium line , wave-length about 2779*6 , is also frequently reversed by a gentle current of hydrogen .
The two aluminium lines near S are frequently reversed when a fragment of the metal is dropped into the crucible , the less refrangible line , wave-length 3091*5 , being more strongly reversed , and continuing reversed for a longer time than that at wave-length 3080'5 .
6*260 5038 6257 both lines of pair .
5036 4535*5 4533*2 4531*7 4690*5 4666*5 4655*5 4299*5 4290*7

