210 PRESENT PROBLEMS <>K IN(>R(}ANIO CHEMISTRY. 



iiiid (IclcnniiH'd hy dciisily. Stsis oUliiiiicd the miinlxT 1 1. 01 (0=10), 

 :iii(l lvicli;ii-(ls has rccciilly (•oiilinncd Ids rcsidls, whdc liaylci^^h and 

 licdiic coiisislciit ly ohiaiiicd densities which, e\('ii when corrected so 

 as to e(|iiali/e the iinniheis of moh'cides in e(|nal volumes, <2ive the 

 l()\v(^i' iioiii-e 14.()()l!. The diU'ei'encc^ is 1 in 1550 — far beyond any pos- 

 sible experimental error. IJecentiy an attempt to combine the two 

 '.nethods lias led to a mean number, but that result can liai-diy be 

 tal<en as (inal. AVhat is the I'eason of the disciv|)an('V? Its disco\'ery 

 will surely advance knowle(l<i;e matei'ialiy. I would su^^'est the 

 preparation of pure compounds of niti'o<:('ii. such as salts of hydi-a- 

 zine, methylanine, etc., and their caicd'ul analysis, and also the 

 accuratx! detei-mination of the density and analysis of such gaseous 

 com|)oimds of nitr(><2;en as niti'ic oxide and jxM'oxide. I have just 

 heai'd from my foi-mer student, W. Iv. W. (Jray, that he has recovei'ed 

 Stas's numbei- by combining 2N() with ()., while the density of NO 

 leads to the lower value for the atonuc weiehl of nili-oe-en. 



'I'he (luestion of the atonuc Avei<j,'ht of tellurium appears to IH^ set- 

 tled, at least so far as its position with reuard to the ec-iierally 

 accepted atomic weifjlit of iodine is concerned. Recent deteiinina- 

 tions <i^ive the fi^jjures 127.5 ((Jutbier), 127.('» (Pellini), and 12V.1) 

 (Kothner). But is that of iodine as accurately known ^ It would 

 a|)pear advisable to revise the determination of Stas, preparing the 

 iodine pi'efei'ably from an oi'gauu' compound, such as iodoform, 

 which can be pi-oduced in a high state of purity. The heteromor- 

 |)hism ol" selenates and tellurati^s, too, has recently been demonst I'ated, 

 and it may be (juestioned whether these elements should l)oth Ixdong 

 to t he same group. 



'V\h\ rai'e eai'ths still remain a pu/zle. Theii- number is increas- 

 ing yearly, and their claim to iudixiduality admits of less and less 

 dispute. What is to be done with them^ Are they to be gi-oupc^d by 

 themselves as Hi'auner and Steele pro|)()se^ If so, how is their con- 

 nection with the other elements to be explained i Recent expei'iments 

 in my laboratory haxc convinced me that in the case of thoi'imn, at 

 least, ordinary tests of |)urity, such as line crystals, constant subliming 

 {)oint, etc., do not always indicate homogeneity; oi- else that we are 

 sadly in want ol" some analyt ical method of sullicient accuracy. The 

 change of thorium into thorium X is pei'ha|)s hai'(ll\' an ex|)lanation 

 ol" the (li\'ei"g<Micies; yet it nnisl be considei'ed ; but cd' this, more anon. 



To tui'U next to another problem closely related to the orderly 

 ai'i'angement of the elements, that of \alency. but little pi'ogress can 

 be chronicled. The suggestions which base been made ai'e si)ecida- 

 tive rather than based on ex])eriment. The existence of many peroxi- 

 dized substances, such as j)ercarbonates, ])ei'b()rates, persulphates and 

 of crystalline compounds of salts with hydrogen peroxide, makes it 

 difficult to draw any indisputable conclusions as regards \alency from 



