the road leading to Pedra Pintada is a place known as Lageiro, where an excavation was made in the rock basin very similar to that already described and for similar purposes. Here again I found many fragments of bones of mastodons and of other mammals, some of them of enormous size, and others smaller than those of the mastodon. On the road leading from Lagoa da Lagea to Aguas Bellas at a place known as Lagoa Cavada is another excavation from which it is said many fossil bones were taken. An examination of the soil thrown out at this place failed to disclose any such fossils however. Large fossil bones were reported to me to have been found at Meirús, a small town three leagues northeast of Pão d' Assucar. I did not see either the excavation or the fossils at this place, but I learned from several intelligent gentlemen that while excavations were being made several years before for making a watering pond for cattle, bones many times as large as those of an ox were found in the soil. It was reported that when this fact became known the government sent persons to preserve these fossils and that the best of them were sent awaywhere I was unable to learn. In every instance that came to my attention the fossils have been taken from what were originally ponds, pools or marshes of fresh water that have subsequently silted up and have been discovered in the digging of artificial reservoirs. The region in which these large fossils occur so abundantly is one now subject to prolonged drouths, and I am disposed to think that the circumstances under which these bones occur suggest, at least, that the animals died of thirst. To any one visiting the drouth regions of Brazil between August and January this explanation seems to be a very natural one. The whole country is parched save the narrow strips of gradually disappearing green along the water courses. The country is but thinly populated and the chief industry is cattle-raising. Beyond these green belts one may travel for many a league without finding a sign of water. When, as often happens, the dry season is prolonged beyond its normal length, the suffering of the cattle is extreme. They are obliged to eat the pulp of the cacti that grow throughout this region, and the herdsmen obtain water for them by digging holes in the sand of the dry river beds. In long drouths, especially when they last for an entire year or even for several years, this source of water supply fails and the cattle must either be driven toward the coast, where water may be had, or they must be left to perish. Under the protection of man cattle may now survive the longest drouths, but left to themselves it is doubtful whether they The channel of the Rio São Francisco seen from the sugar loaf hill at Pão d'Assucar, State of Alagoas; looking up stream. could withstand the long drouths that occasionally occur in the interior of northeast Brazil. I was led to this conclusion when I first visited the interior of Alagoas and Pernambuco in 1876. I quote from Dr. J. W. Gregory's "Great Rift Valley" what seems to be corroborative evidence that in Africa large animals do perish as I have suggested:* "Here and there around a water-hole we found acres of ground white with the bones of rhinoceros and zebra, gazelle and antelope, jackal and hyena, and among them we once observed the remains of a lion. All the bones of the skeletons were there, and they were fresh and ungnawed. The explanation is simple. The year before there had been a drouth which had cleared both game and people from the district. Those which did not migrate crowded around the dwindling pools, and fought for the last drop of water. These accumulations of bones were due, therefore, to a drouth and not to a deluge." It is much to be desired that the fossil Pleistocene mammals of Brazil be studied systematically. The work of Lund upon the cave fauna of Brazil is classic, but no attempt has yet been made to get together the material from the ancient watering places of northeast Brazil, while the fragmentary collections in the Musen Nacional and in the hands of the other scientific organizations and of private individuals are undescribed and unknown. The only publication that has been made upon the subject is a paper by Dr. F. L. C. Burlamaqui published at Rio de Janeiro in 1855 by the Sociedade Vellosiana entitled "Noticia acerca dos animaes de raças extinctas descobertos em varios postos do Brasil." Stanford University, California, December 10, 1901. *The Great Rift Valley, by J. W. Gregory, London, 1896, pp. 268-269. ART. XII.-The Estimation of Copper as Cuprous Sulphocyanide in the Presence of Bismuth, Antimony, Tin and Arsenic; by R. G. VAN NAME. [Contributions from the Kent Chemical Laboratory of Yale University, -CVI.] IT has long been known that copper may be quantitatively separated from a number of other metals by precipitation as cuprous sulphocyanide. The additional fact that this salt is suitable for direct weighing as a means of estimating copper was mentioned, apparently for the first time by Rivot* in 1854, and was confirmed a quarter of a century later by the experiments of Busse, who employed the process for determining copper not only alone but also in the presence of iron, nickel, zinc and arsenic. Recent work by the present writert has shown that, with the sole modification of filtering and weighing upon asbestos, Rivot's method is accurate and satisfactory. In view of the tendency of bismuth, antimony and tin to form insoluble basic chlorides in solutions containing only a slight excess of hydrochloric acid, and the difficulty of obtaining a complete precipitation of cuprous sulphocyanide in strongly acid solutions, the estimation of copper by the above method in the presence of these metals seemed to present some difficulties. The following investigation, in which arsenic was also included on account of its close relationship with the above metals, was accordingly undertaken. The first series of experiments yielded the results given in Table I, the method being as follows: A convenient quantity of a copper sulphate solution of about normal strength was measured from a burette, diluted, and acidified with a definite amount of hydrochloric or sulphuric acid. If tartaric acid was employed it was added at this point, after which the solu tion of arsenic, bismuth or antimony was introduced and the copper finally precipitated by ammonium bisulphite and ammonium sulphocyanide. Of the last a uniform amount, 70cm* of a decinormal solution, was used for each experiment. The ammonium bisulphite solution was prepared in the usual way by passing sulphur dioxide into strong ammonia. Arsenic was introduced in the form of measured portions of a standard arsenious acid solution which were made faintly acid to litmus just before addition. Bismuth and antimony were taken in the form of solutions of the chlorides in dilute hydrochloric acid, the amount of acid thus introduced into. the determination being in all cases taken into account. * Compt. Rend., xxxviii, 868. Zeitschr. anal. Chem., xvii, 53. |