November 1, 1888.] 



♦ KNOWLEDGE ♦ 



17 



o'clock in the morning and placed in an empty ward. He 

 shortly became unconscious, and died the same evening. 

 It is a question for consideration whether the deaths 

 brought about by the official precautions adopted during 

 yellow fever and cholera .scares do not outnumber the deaths 

 due to the dreaded scourges themselves. In the particular 

 instance under consideration, it does not appear certain that 

 if the case had been really one of yellow fever, the danger of 

 infection would have been diminished by his removal 

 down the hotel stairs and through the streets. The report 

 states that — 



" The only positive change due to disease which could be 

 made out was in the kidneys, which showed the appearance 

 of old, though not advanced, disease. The alterations pro- 

 duced in the body by yellow fever are usually of such a 

 character as to be neaily or completely obliterated by such 

 advanced decomposition as the organs presented. We are, 

 therefore, only able to say in this connection that there was 

 no other evident cau.se of death, and nothing which would be 

 incompatible with death from that disease. The final con- 

 clusion as to the cause of death must, therefore, in our 

 opinion, be largely bastd upon the clinical his-tory. 



T. Mitchell Pridden, M.D., 

 Hermann M. Biufitis, M.D., 

 Pathologists to the Health Department 

 of New York City. 

 * * * 



Admiral Mouchez, the Director of the Paris Observa- 

 tory, where experiments are now being made with regard 

 to the plates to be used for the photographic charts of the 

 heavens, reports that he has received from America some 

 plates which are superior in sensitiveness to any that have 

 been previously employed. The American plates have been 

 tried by the MM. Henry, who report on them most favour- 

 ably as to the uniformity of film and extreme sensitiveness. 

 They have been able to obtain with one hour's exposure 

 results which were obtained with difficulty in two hours 

 with the best European plates. Tliis is most satisfactory, 

 as the great length of exposure required for fainter stars is 

 one of the chief difficulties in the way of the execution of 

 the photographic chart . 



* * * 



The electro-melting process for making aluminium 

 alloys may exercise a considerable influence on the arts in 

 the future. For, according to a paper read before the 

 Franklin Institute by Mr. Pemberton, there is hardly any 

 metal with which aluminium will not alloy, and it alloys 

 ■with nothing that it does not improve. Aluminium bronze 

 is the strongest of the cast metals, and, when rolled, the 

 .strongest of i oiled metals. It does not corrode, as ordinary 

 bronze does, and its fine golden colour gives it a pleasing 

 appearance. Aluminium brass is noted for its gieat strength 

 and high elastic limits. Aluminium iron is now extensively 

 used for castings, for the fusing point of wrought-iron is 

 reduced almost 500° by the addition of only one- thousandth 

 of its weight of aluminium. 



ANOTHER UNDECIPHERABLE CIPHER. 



(To the Editor f*/" Knowledge.) 

 WAS much interested in your cipher in the 

 July number. It is no doubt quite unde- 

 ii|ihcialilo in the absence of the key, being 

 busi'd upon the principle of fluctuating signs, 

 which is the only practical principle for a real 

 cryptogram. That is to say, that A shall not 

 alw.ays be written down as .v, but also as 

 Your method of obtaining the fluctuations is 



ingenious and comparatively easy, much more so than the 

 old " Chiffre indechiffrable " 

 abed 

 be de 



c d e f (tc, (tc, 

 used with a prearranged key-word. I doubt, however, 

 whether you gain much by the knuztheloky part of 

 your system. Is there any real need to indicate the number 

 of letters in each word 1 Instead of recommencing 2,01.5, 

 &c., at the beginning of each word, you might run the 

 20,153,401, ifec, right through the message. After a crypto- 

 gram has been translated back into a message, the latter is 

 quite legible even though the divisions between the words 

 be not marked. If my suggestion were approved of, you 

 would sire writing down the number of letters in each 

 word (2, 3, 8, 2, 3) on your paper of preparation, and also 

 save writing down the corresponding key-word {nutnuy) at 

 the commencement of the cryptogram. 



I also have invented an " Undecipherable Cipher." It is 

 on the piinciple of fluctuating signs, but I obtain the 

 fluctuations by quite a different plan to yours, and I should 

 be glad if you will allow me to submit it to a trial by 

 experts in the columns of your paper. 



Take any six letters of the alphabet, say, for example, 

 aco s u V. Make them into combinations of twos, thus, a a, 

 a c, a 0, lie, in all thirty-six combinations. These will be 

 our signs. Distribute the letters of the alphabet, abode, 

 &c., in any order whatever, amongst the above combina- 

 tions. This will absorb twenty-six of the combinations. 

 Amongst the remaining ten I distribute an additional letter 

 e, the words And, The, You, the " Letter Piepeat,"' and three 

 " End of Word." The use of these will be explained below. 



Our code will now stand thus, for example : — 



Three or four codes may be prepared, all of coui-se diflerent. 



The first rule for writing the cryptogram is to write all 

 words continnou.sly — that is, without any spaces between 

 them. Thus " We are " would appear as sacooscvss. So far, 

 however, the combinations of two letters would in rejility 

 have only the same effect as single chai-.ictei-s. An enemy, 

 or investigator, by commencing at either end, and dividing 

 the letters into groups of two, would have pnvctirally a 

 certain number of single signs, with which he could deal 

 further after the usual analytical methods employed in 

 decipherment. 



I came now to the cardinal point and distinguishing 

 feature of my system. At the end of each word the double 

 sign (au, or oc, or .in) is written to denote the end of the 

 word, and, after the dovible sign, a sim/le extra false letter 

 (in Code No. 3 either a, c, o, s, u, r) is inserted. T/ie effect 

 of this single fahe letter is virtually to c/mnge all the succeed- 



