Limnophis crassus, gen. et sp. nov. One of the most interesting of the Ophidian fossils obtained during our explorations in Wyoming was a single anterior dorsal vertebra, very well preserved, and quite unlike any hitherto described. It indicates a moderately sized constricting serpent, perhaps six feet in length, but evidently of a very different type from the species already characterized. On comparing it with the corresponding vertebræ of Boavus and Lithophis, a striking difference is at once seen in the dimensions of the articular cup, which considerably exceeds in width the parallel diameter of the base of the zygosphene, a feature only observed heretofore in a few fossil serpents from the Eocene, and quite unknown in modern species. The cup and ball have a subtriangular, ovate outline, the greatest transverse diameter being above the center: they are also placed nearly at right angles to the axis of the centrum. Another marked peculiarity of this vertebra is the unusual posterior extension of the neurapophyses, which project some distance beyond the articular ball. Their nearly flat sides slope downward and outward like a gable roof, without any indication of the angle so characteristic of the vertebræ of Palæophis, which in several other respects the present specimen strongly resembles. The neural spine is unfortunately not preserved, but it was evidently short, and confined to the posterior two-thirds of the neural arch. The anterior zygapophyses had only a moderate expansion. The diapophyses have, for the attachment of the rib, a single, rounded, and prominent tubercle, with its lower border above the inferior margin of the cup, and with no indication of a dependant process. There is also no ridge extending from the diapophysis to the articular ball. The hypapophysis consists of a single, compressed, obtuse tubercle, which descends from the middle of the centrum downward and slightly forward; its base occupying rather more than half of the median line. The more important dimensions of this vertebra are as follows: Length of centrum from edge of cup to convexity of ball, 3.20 lines. Transverse diameter of cup,. Vertical diameter of cup, Vertical diameter of ball, Distance from posterior summit of neurapophyses to inferior margin of ball, Length of hypapophsis below inferior margin of cup,- 2.30 This interesting vertebra, although the only one of the kind at present known, is so characteristic, and well preserved, and differs so widely from the corresponding remains of Ophidians *Memoirs of Paleontological Society of London, 1850, p 57. hitherto described, as to justify establishing for its reception a new genus, which may be called Limnophis.* The species may properly be named Limnophis crassus, in reference to the probable proportions of the extinct serpent thus indicated. The specimen was discovered by Mr. A. H. Ewing, a member of the Yale party, in the Eocene fresh-water deposits, near Marsh's Fork, about fifteen miles from Fort Bridger, in Wyoming. Yale College, New Haven, March 20th, 1871. ART. XLIX.-Contributions to Chemistry from the Laboratory of the Lawrence Scientific School. No. 14.-On the Estimation of Phosphoric Acid; by CHARLES E. MUNROE. 1. THOUGH the methods commonly in use for the quantitative determination of phosphoric acid are capable in most cases of giving accurate results, yet it will, I think, hardly be claimed for them, that they render new methods unnecessary, and I have therefore undertaken the following investigation in the hope of adding something of value to our analytical resources. My attention was first called to the employment of ferric chloride with the addition of sufficient mercuric oxide to neutralize the excess of acid. Free mercuric oxide was first added directly, but, as a complete separation could not be effected in this manner, it was found more advantageous to produce the oxide in the solution itself. This was accomplished by first adding mercuric chloride and then potassic or sodic hydrate. As sodic hydratet can be obtained in commerce absolutely pure this was employed. The precipitate, consisting of ferric phosphate and ferric and mercuric oxides was then evaporated to dryness in the manner recommended by Chatard,+ filtered and washed on the Bunsen pump.§ Unhappily the phosphoric acid redissolved after evaporation to dryness and the process was therefore abandoned. Bunsen's delicate test with magnesium wire was used throughout the work, sometimes in connection with ammonic molybdate, for the detection of phosphoric acid. * Aíuvn, lake, and "Opis, serpent. The article used was manufactured in England by the direct oxidation of metallic sodium. This Journal, vol. iv, p. 247. A simple modification of the Bunsen pump may be obtained by connecting a piece of glass tubing, by means of rubber, with a Bunsen funnel. A funnel holding 50 cm. connected with a piece of glass tubing one meter in length emptied itself five times while another funnel of the same capacity, arranged as usual, did so but once. The pressure was so great that the filter had to be protected with a platinum point to prevent it from giving way. AM. JOUR. SCI-THIRD SERIES, VOL. I, No. 5.-MAY, 1871. I next studied the behavior of aluminic oxide toward phosphoric acid, and this time met with better success. The following course was pursued. To the boiling phosphate solution a weighed quantity of pure aluminic sulphate, previously dissolved, was added. A solution of mercuric chloride was then added, and finally sodic hydrate until a precipitate of mercuric oxide was obtained which remained undissolved. In order to hasten the operation the precipitate was allowed to settle and the supernatant liquid poured upon the filter. The gelatinous precipitate was then evaporated to complete dryness as before mentioned, filtered, ignited and weighed. The increase of weight, over that of the aluminic oxide used, was phosphoric oxide. It was found extremely difficult to burn the filter. Care must be taken that for every gram of phosphoric oxide at least two grams of aluminic oxide are added. The percentage of aluminic oxide, in the sulphate used, was determined by ignition, ammonic carbonate being used to drive off the last traces of sulphuric oxide. The following results were obtained : 66 66 66 (1) 7320 grufs. Al2(SO4)3. 18H2O gave 1310 grms. Al2O, 17.89 p. c. Anhydrous disodic phosphate was the first salt treated. In order to work with greater rapidity a solution of known strength was prepared of which 50 c.m., corresponded to 6195 grms. of phosphate. This gave the following results: Ammonio-sodic phosphate (Na(NH)H Po, 4H,O) was next. analyzed. The amount of phosphoric oxide in the salt was found by igniting it and weighing the sodic metaphosphate. (1) 1.4650 grms. gave 6172 grms. NaPO, =29-32 p. c. P,Os (2) 1·5017 ⚫6344 66 66 =29-40 By the alumina process the following results were then obtained. Calcic phosphate did not give good results by this process on account of the difficulty of washing out the calcic sulphate formed. I succeeded no better with ammonio-magnesic phosphate doubtless because some of the magnesia went down with the alumina. These experiments are sufficient to show that phosphoric acid may be determined with extreme precision in the alkaline phosphates by the method above given, but the method applies with advantage only to these salts. On the other hand it must be remembered that many phosphates may be decomposed by fusion with the alkaline carbonates and the phosphoric acid then determined as above, so that in fact the method applies in perhaps the greater number of cases which occur in practice. § 2. Rose's* method for the separation of phosphoric acid by means of mercurous nitrate and subsequent estimation in the form of pyrophosphate of magnesia has been much used. It will, however, be acknowledged that it has some serious defects. It is tedious, the many operations required increase the chances of error, and the fusion of mercurous phosphate with the mixed carbonates exerts, according to Rose, a sensible action upon the platinum crucible. Nevertheless the perfect separation which it gives is a strong recommendation of the method. It appeared possible that a valuable modification of this process might result from the addition of mercuric oxide in the manner already employed in the alumina process above described. The experiments were executed in the following manner. To a boiling solution of the phosphate mercurous nitrate was added in slight excess. This threw down a lemon-yellow colored, crystalline precipitate of nitro-mercurous phosphate which adhered to the sides of the beaker when touched by the stirring rod. Mercuric nitrate was then added and finally sodic hydrate until a slight precipitate of mercuric oxide was obtained which remained undissolved. The mixture of the two salts was found to be insoluble in both hot and cold water, and to wash like sand. The next step was to find a readier means of treating the nitro-mercurous phosphate. It was first proposed to dissolve the washed and dried precipitate in chlorhydric acid and titrate with potassic permanganate so as to oxidize the mercurous to the mercuric salt. This process failed. The salt was then treated with sulphuric acid. The final reaction was perfectly sharp, but the results were not satisfactory. These results having proved that nothing was to be gained in that direction, I tried to remove the mercury by decomposing the nitro-mercurous phosphate with nitric acid and precipitating the mercury with sulphydric acid or ammonic sulphide. The phosphoric acid was then to be determined in the filtrate by the ammonio-magnesian solution. On examining the mercurous Traité de Chimie Analytique, ii, 708. sulphide it was found that some of the phosphoric acid was always precipitated with it. Its presence was proved by Bunsen's test with the magnesium wire. The salt was then dissolved in a solution of potassic cyanide and a current of sulphydric acid passed through, but with the same result as before. Again it was thought that the phosphoric acid might be precipitated directly from the cyanide solution by ammoniomagnesic chloride, but experiment showed that the cyanide contained so much cyanate and carbonate that, although the solution was boiled and chlorhydric acid was added to it, accurate results could not be obtained. As the ignition of the nitro-mercurous phosphate with the non volatile metallic oxides, in order to drive off the mercury, promised well, this process was then tried in the following manner. The dried salt was thoroughly mixed with a weighed quantity of cupric oxide in a platinum crucible and the filter placed on top. The whole was ignited, at first gently, then to low redness, at the mouth of a muffle,* then cooled, a few drops of nitric acid added to oxidize the cupric oxide reduced by the filter, and then re-ignited. The ignition was continued until the weight became constant. The increased weight was the phosphoric oxide (P,O,). After ignition the cupric oxide and phosphate came out of the crucible in a beautiful coherent mass, leaving it perfectly clean and unharmed. Disodic Phosphate. It is quite difficult to obtain this salt anhydrous. It becomes so between 30° and 40°, but, as it fuses at 35°, it must be repulverized for analysis, and as it is quite hygroscopic, it is liable to change during the operation. A quantity dried over sulphuric acid was used, the phosphoric oxide being determined by ignition. 66 66 •5075 66 2 per. ct. = 21.97 2 5 66 (1) 11130 grms. gave 4576 grms. Na P2O, = 21-98 P2Os (2) 1.2355 The mercurous oxide process with cupric oxide gave the following results. Found. Theory. (1) 10245 grms. gave 5124 grms. P2O 2 5 (2) i'2198 4716 66 Number one was a specimen of anhydrous phosphate. It will be noticed that the second analysis is much too high. In * Justice is not done the muffle furnace. It gives an even, mellow heat which can be easily regulated, and for igniting precipitates it is unequaled. I have before spoken of the difficulty found in burning the filter with the aluminic phosphate. By placing it at the mouth of the muffle I was able to ignite it in the most perfect manner, without tipping the crucible, in less than ten minutes. Discretion must be used in placing the crucible in the muffle. |