72 PROCEEDINGS OF THE AMERICAN ACADEMY 



been placed so high as 39°, or about that of the iso-hydride itself. 

 In the absence of evidence, therefore, that the body described by 

 Frankland boiling at 30° is not of the series of hydrides, it seems 

 reasonable to take this, the one first described and best studied, as 

 typical of this series ; and to consider, for the present, those bodies 

 that correspond with this in regard to their boiling-points and 

 other properties, viz. those boiling at 0°, 60°, 90°, and 120°, re- 

 spectively, as also of the same series; and those boiling at 9°, 39°, 

 69°, and 129°, as a series of isomers, or iso-hydrides ; for it is well 

 known with regard to the best-studied of homologous series of liquid 

 bodies that, as a rule, a common boiling-point difference between 

 the contiguous members prevails throughout any given series; any 

 deviation from this being so exceptional as to justify a suspicion 

 of error.* Nevertheless, the series of hydrides is generally repre- 

 sented with some of the boiling-points of one series and some of 

 the other, while others do not agree with the true boiling-points 

 of either series. It is hoped that the results recorded in this paper 

 may serve to bring order out of this confusion. 



Analysis.^ 



0.1935 grm. of the substance gave 0.5912 of carbonic acid, and 

 0.286G of water. 



* Note Goldstein's late paper. 



t The analyses given in this paper having been made, for the most part, pre- 

 viously to the working out of my process for combustion in oxygen gas, above 

 referred to, it is to be understood, unless otherwise stated, that they were made, 

 so far as relates to those boiling under 150°, by distilling the vapor of the sub- 

 stance into a column of ignited oxide of copper, and completing the combus- 

 tion in a stream of oxygen. This method, however, was not at first successful. 

 So long as I continued to follow the direction of Regnault to connect the bulb 

 with the combustion tube by means of a caoutchouc tube, I was unable to 

 avoid a loss of several per cent, occasioned, doubtless, by the escape of sub- 

 stance through the material of the connecting tube. That this was the proba- 

 ble source of error was shown by suspending a piece of this tubing in the vapor 

 of petroleum naphtha at the common temperature, the walls of the tube soon 

 acquiring two or three times their original thickness, from absorption of hydro- 

 carbon vapor. During an analysis, however, no change in the appearance of 

 the tubing was noticed, — any perceptible thickening being prevented by the 

 heat and exposure to air carrying away the vapor as fast as absorbed. To get 

 over this difficulty, I contrived a convenient method of making the connection 

 with the combustion tube by means of a cork. A tube, 4 or 5 inches or more 

 in length, with a small bulb blown in the centre, and with capillary ends, was 

 substituted for the ordinary bulb. By leaving the tube of sufficient length on 

 each side of the bulb, the same bulb could be repeatedly used, only requiring 



