LIBRARY OF THE UNIVERSITY OF CALIFORNIA. RECEIVED BY EXCHANGE Class An Electrolytic Method of Analyzing Zinc Ores THESIS Presented to the Faculty of the Department of Philosophy of the University of Pennsylvania in Partial Fulfilment of the Requirements for the Degree of Doctor of Philosophy, BY HARRISON HALE SPRINGFIELD, Mo. 1908 HODGES & KIRK PHH.ADEI.PHIA 1908 An Electrolytic Method of Analyzing Zinc Ores THESIS Presented to the Faculty of the Department of Philosophy of the University of Pennsylvania in Partial Fulfilment of the Requirements for the Degree of Doctor of Philosophy, HODGES & KIRK PHILADELPHIA 1908 This work was done at the suggestion of Dr. Edgar F. Smith, and the writer wishes to express most sincere thanks both for the assistance given in this investigation and for unfailing personal kindness. INTRODUCTION. The application of electro-analysis with its accuracy, its clean- liness and its ease of execution to the determination of mineral constituents has obvious advantages. This has long been recog- nized. More than thirty years ago Parodi and Mascazzini (i) published a method for the determination of zinc by adding an excess of ammonium acetate to a solution of its sulphate; this method was recommended for use in the analysis of zinc ores. In 1902 Edgar F. Smith (2) after expressing confidence that electrolytic methods would prove "wholly satisfactory" on ap- plication to natural products gives results of the determination of zinc in a pure blende, using a sodium acetate electrolyte, the zinc being present as sulphate. The introduction of the rotating anode with zinc solutions by Exner (3), greatly increasing the advantages of the electrolytic method has made an imperative de- mand for further work along this line. With these facts in view a study of zinc blende was undertaken with the purpose of obtaining not only a method for the determ- ination of the metal but also of the sulphur content of the min- eral. The results have shown that both may be determined with ease from the same sample and in a shorter time than by any other satisfactory process. Attention has also been given to the form of electrodes, es- pecially the cathode, best suited for this determination. Further work was then done to test the general applicability of the method for determining the metal from any zinc ore with fair success. Such a practical method is particularly desirable on account of the present unsatisfactory condition of rapid zinc estimation. (i). Ber., 10, 1098. (2). J. Am. Ch. S. 24, 1073- (3). J. Am. Ch. S. 25, 899- 186924 PRELIMINARY WORK. A pure zinc blende finely ground was used to try out the method as follows : A weighed portion of the ore was placed on about fifteen grams of sodium hydroxide which had been pre- viously fused in a nickel crucible and allowed to cool. This was then heated to quiet fusion, the top of the Bunsen flame touching the crucible, for fifteen or twenty minutes. The heat should be strong enough to completely decompose the ore but the fusion should not boil. On cooling, the melt was taken up in water, filtered, diluted to 125 c. c. and electrolyzed in a silver- coated platinum dish with a current of N. D. 100 = 5 amperes and 4 volts, using a flat spiral anode of platinum wire (about i mm. diameter), making approximately 600 revolutions per minute. The glass covers were washed down with hot water five to ten minutes before the current was interrupted. The liquid was siphoned off without breaking the current, water being poured in, The deposit was further washed with water then with abso- lute alcohol and ether. (See Smith's Electro- Analysis 4th edi- tion, p. 116). This gave: Number Zinc blende Zinc found Zinc gram gram per cent. 1 0.2378 0.1549 65.14 2 0.3842 0.2505 65.20 3 0.3790 0.2474 65.28 4 0.2733 0.1831 67.00 In No. 4 an effort was made to determine zinc without filter- ing simply taking up the fusion in the crucible with water, decanting into a dish and electrolyzing. The high result was due to impurities in the deposit, iron being found and silica also probably being present. The zinc was removed from the dish by dilute sulphuric acid (i : 50) which had almost no effect on the silver coating. GENERAL EXPERIMENTAL WORK. For this work a larger quantity of another sample of pure blende from Joplin (i) was ground so as to pass through a 120 mesh sieve. The method was the same as in the preliminary work. However potassium hydroxide (Kahlbaum's sticks pure by alcohol) was used for the fusion, it being considered purer. The fusion was taken up in water and filtered through a Gooch crucible fastened by a stopper in the top of a bell jar, into a nickel dish. The work of Mr. J. S. Goldbaum in this laboratory had shown this to give better zinc deposits and to be more easily cleaned than the silver-coated platinum dish. The deposition was more rapid if the solution was warmed before the current was started. A. Nickel dish cathode, hemispherical, 5cm. diameter. Solu- tion diluted till the cathode surface was 130140 square cm. Flat spiral anode of platinum wire as in preliminary work. These results were obtained: Number Zinc blende gram 0.4513 0.6520 0.7279 0.5640 0.4242 0.3871 0.5557 0.4686 0.6325 0.3914 0.4033 0.4058 0.4852 0.4264 0.4226 9 10 ii 12 13 14 3 II 19 Time heated Current Zinc found Zinc minutes Volts Amperes Time gram per cent. 30 So 30 60 35 4p 30 Average 15 determinations 64.87 4 5 ; 30 0.2937 65.08 3-5 0.4215 64.65 5 45 0.4766 65.48 > < 30 0.3622 64.22 < 0.2732 64.40 ti t 50 0.2507 64.76 ( ( 0.3601 64.80 M 40 0.3024 64-53 5-5 60 40 0.4108 0.2531 64-94 64.66 4-5 50 0.2621 64.99 0.2658 65.50 30 0.3133 64.57 M 50 0.2764 64.82 ( 1 40 0.2772 65.59 In No. 14 to remove crystals of potassium sulphate held up by the zinc (due to addition of sulphuric acid and more potas- sium hydroxide to remedy a poor fusion) and in Nos. 17 and 19 to remove silica (due to imperfect filtration), the zinc deposit was dissolved in dilute sulphuric acid (i : 50), an excess of potas- sium hydroxide added and the solution again electrolyzed. In these determinations it was noted that a rapid rotation of the anode, about 600 revolutions per minute, greatly improves the character of the zinc deposit, preventing sponginess and making it adherent. With this precaution beautiful deposits were generally obtained, smooth and adherent. When the elec- trolysis is first begun there is considerable frothing which grad- ually diminishes, becoming least when most of the zinc is de- posited. During these and subsequent determinations no attempt was made to reduce the time to a minimum, the aim being to first test the efficiency of the process. After the melt is in quiet fusion almost no attention is required for it. Save for washing down the cover glasses the operator need give practically no time to the electrolysis after turning on the current. The writer is confi- dent that fifteen minutes of quiet fusion is sufficient for half a gram of ore and fifteen to thirty minutes for the deposition of the zinc, varying with the content of the ore and the volume of electrolyte. Tests. The residue on the Gooch crucible was frequently tested for zinc with cobalt nitrate in the oxidizing flame before the blowpipe (Rinmann's Green), and no zinc was found in the determinations given in this paper. The siphonate was always tested by acidifying with hydrochloric acid and adding potassium ferrocyanide with negative results for zinc in the determinations given above. Raising the level of the liquid and noting whether there is a further deposition of the metal is an aid in determining when the reaction is complete; but this becomes unnecessary when a number of determinations are being made as the analyst soon learns the length of time required. The zinc deposit was dissolved off the dish in dilute sulphuric acid, a few drops of nitric acid added and the solution boiled. On cooling, this was tested for iron with potassium ferrocyanide or sulphocyanide. No iron was present with the zinc deposit in any of the determinations previously given. When the fusion was not sufficiently heated a cloudy purplish mixture was obtained on adding water. This substance was hard to filter and invariably left zinc in the residue on the Gooch crucible. This exceptional behavior is due to two causes. First : at the lower temperature the iron was not completely changed to the insoluble ferric oxide, and some potassium ferrate was formed. This gives the color which is lost on making the mix- ture acid. Second: under the same conditions the sulphur was only partially oxidized, giving some zinc suphide, sulphite and thiosulphate which remained on the filter causing the test men- tioned. This is proven t>y the addition of hydrochloric acid in excess of the potassium hydroxide, when hydrogen sulphide was first given off and then sulphur dioxide with a copious separa- tion of sulphur. These difficulties are all avoided on heating the fusion to a higher temperature not to boiling, but so that the bottom of crucible is touched by the top of the Bunsen flame. This causes the formation of the insoluble ferric oxide and gives no insoluble zinc compounds on the Gooch crucible. An attempt was then made to avoid filtration before the elec- trolysis. Experiment No. 4 in the preliminary work had shown that this could not be done in a dish. Consequently a cylindrical nickel-plated platinum gauze was used as cathode, and as anode a small spiral of platinum wire, both being suspended vertically in a beaker, the anode rotating at the rate of at least 500 revolu- tions per minute. It was thought that ferric oxide, silica and other solid matter would drop to the bottom of the beaker and not adhere to the cathode. Six years before Smith (4) adopted a similar pro- cedure in electrolyzing an ammoniacal copper solution containing ferric hydrate from chalcopyrite. He found "that if platinum dishes be employed as cathodes to receive the copper, the latter has invariably been contaminated with iron, but that if platinum cylinders, dipping into the ammoniacal solution contained in beaker-glasses, were used for cathodes the copper deposited upon them was free from iron." The fusion was taken up in water and poured directly into the beaker and after warming was analyzed with these results: B. Nickel-plated platinum gauze cathode vertical 25 wires per cm. of 0.2 mm. diameter, 10 cm. long, 5 cm. wide, giving a cathode surface of approximately 150 square cm. A stout plati- num wire is split for a part of its length and riveted over the center of the gauze, the upper end being used to make the elec- trical connections and bent twice at right angles to be clear of the rotating anode. The gauze was bent into cylindrical form, but the cylinder was not entirely closed, the distance between the (4). J. Am. Ch. S. 24, 1076. ends being one to two cm. As anode a piece of platinum wire, about i mm. diameter, which had been wound into a spiral about an ordinary lead pencil vertical. Number Zinc blende Time heated Current Zinc found Zinc gram minutes Volts Amperes Time gram per cent. 20 0.5064 30 4.5 5 55 0.3275 64.67 " ' 50 0.3127 65.40 21 0.4720 22 0.5526 30 0.3612 65.36 23 0.4752 40 55 0.3066 64.52 24 0.4298 45 0.2804 65.24 25 0.4235 50 0.2735 64.58 26 0.4826 5 60 0.3146 65.18 37 0.4454 90 5.5 4.5 45 0.2900 65.11 65.01 Less iron.. 0.08 Average 8 determinations 64.93 A slight trace of iron was generally found in zinc deposit ; this was estimated colorimetrically with potassium sulphocyanide and standard iron solution to be 0.08%, which correction is made above. The tests for zinc in siphonate and residue resulted negatively. The smallness of the amount of iron in the zinc deposit seems to show that it is held mechanically, being carried by the zinc in its deposition. The agitation of the solution by the rotating anode would be favorable to this. This agitation may account for the difference in the result from the work of Smith on chal- copyrite when the copper deposits contained no iron using sta- tionary vertical electrodes. It was thought that a sheet of metal might offer less oppor- tunity for retaining the iron. Accordingly the following determ- inations were made : C-i. Nickel-plated sheet platinum measuring approximately 4 cm. by 10 cm., giving 80 square cm. of cathode surface ; partly closed cylinder vertical. Same vertical anode as in B. Number Zinc blende Time heated Current Zinc found Zinc gram minutes Volts Amperes Time gram per cent. 27 0.4296 40 5 5 70 0.2797 65.10 28 0.4172 . 40 0.2698 64.66 The nickel showed a tendency to peel off from the platinum so no more determinations were made. One determination was tried, using for the cathode an in- verted nickel crucible, the bottom of which had been cut out to admit the passage of the anode. This was suspended by a nickel wire. C-2. This closed nickel cathode bell-shaped was not satis- factory, the solution not being sufficiently agitated by the rota- tion of the anode. The deposit was almost entirely on the inner surface of the cathode, which did not afford an area large enough to hold all of the zinc. -3. A sheet nickel cathode 9.5 cm. by 5 cm., giving a surface of 95 square cm., of similar shape to that used in C-i vertical; cathode suspended by nickel wire. Same vertical anode as in B. Number 29 30 31 32 33 34 Zinc blende gram 0.4104 0.4132 0.4094 0.5024 0.3681 0.4084 0.4555 0.3946 Time heated Current minutes Volts Amperes Time 40 50 40 90 4-5 5 5-5 40 u 45 40 50 1? inc founc 1 Zinc gram per cent. 0.2678 65.25 0.2669 64.59 0.2652 64.77 0.3291 65.50 0.2398 65.14 0.2665 65-25 0.2964 65.07 0.2579 65.35 65.07 Less iron.. 0.12 Average (C-i, -3) 10 determinations 64.95 A trace of iron found in the zinc deposits, was estimated and corrected for as in B. There was no zinc in siphonate or residue left on filtering. With the sheet nickel cathode the zinc was not always ad- herent, several determinations being lost on this account. It should also be noted that the cathode surface as stated for a sheet metal electrode is not comparable to that for a gauze cathode. Average of 33 determinations giving equal value to each determination 64.91% Average of 3 series of determinations with different forms of electrode, giving equal value to each series 64.92% From the previous experiments it appears preferable in the determination of zinc by fusion with alkali and electrolysis of the solution, to use a vertical gauze cathode of nickel, dipping into the solution in a beaker. This conclusion is based on these con- siderations : 10 1. Rapidity of weighing. The gauze being much lighter than the dish and somewhat lighter than the sheet metal, quickly comes to constant weight on the balance pan, giving a more accurate re- sult with greater speed. 2. Slight interference of gangue or other suspended matter. Indeed, unless the amount is considerable, the interference is neg- ligible. This advantage is greatest over the dish, where the electro- lyte to give the best results must be perfectly clear. It is seen also that the gauze has here a slight advantage over the sheet metal. 3. Large cathode surface in small amount of electrolyte. This advantage is not only in the greater surface of the cathode, but the easy flow of the solution through the gauze gives a more even deposit over its entire surface and makes a given surface of correspondingly greater efficiency than is possible with a sheet of metal of cylindrical form, even when the cylinder is not com- pletely closed. If the cylinder is entirely closed the outer surface plays a still smaller part in receiving the deposit. 4. Rapidity of deposition and adherent character of deposit. The rapidity of the deposition is a result of the increased cathode surface allowing the use of a more concentrated electrolyte, while the gauze offers but little opportunity for the zinc to peel off, which danger is to be guarded against with the other forms of cathode. At the end of the electrolysis the vertical cathode, either gauze or sheet, may be gradually raised from the electrolyte while being washed with cold water, the current not being broken. This procedure is in some cases easier than the usual process of siphoning. Washing with alcohol and ether is, of course, carried out as usual before weighing. Such a course is of decided ad- vantage when the sulphur is to be determined as well as the metal, thereby avoiding a large dilution of the sulphate solution. There is no reason for using gauze of nickel-plated platinum, a nickel gauze serving just as well. The plated nickel electrode was used in this work because it was at hand and a good nickel coating was easily made from an ammoniacal double sul- phate solution of nickel and ammonium, and served for numerous determinations. The cheapness of a nickel cathode is, however, a decided advantage. While it has been shown that determinations of zinc in a pure blende can be made without filtering, using a vertical gauze 11 cathode and correcting slightly for iron, yet it is generally better to filter, and is really a necessity if the amount of insoluble mat- ter is large. Determination of Sulphur. The electrolytes remaining after the deposition of the metal in Nos. 37 and 38 were concentrated, made acid with hydrochloric acid, and after filtration to re- move silica and gangue, barium chloride was added with the usual precautions, giving: Number Zince blende Barium sulphate Sulphur found Sulphur gram gram gram per cent. 37 0.4454 1.0347 0.14206 31.89 38 0.4178 0.9722 0.13348 31.94 A slight trace of sulphate was found in the potassium hy- droxide used in fusion. Gravimetric analysis of this same blende showed : Gangue 1.18% (1.18)% Iron Fe 0.41 o .41 Sulphur S 31.72 31.79 Zinc Zn 64.96 65.10 Calcium carbonate. . .CaCOs 1.55 (i-55) 99.82 100.03 Gangue and calcium carbonate were determinted in only one sample, the latter being calculated from calcium oxide. The ore was carefully oxidized by dilute nitric acid; this was re- placed by hydrochloric acid, and after filtering out the gangue and later the ferric hydroxide precipitated by ammonium hy- droxide, the zinc was precipitated as sulphide. This zinc sul- phide was dissolved with dilute nitric acid from off filter into weighed porcelain crucible, evaporated to dryness on water bath and ignited with blast to constant weight as zinc oxide. A comparison of methods shows : Zinc (percentages) Sulphur Electrolytic Gravimetric From siphonate Oxidized by nitric acid 64.87 64.96 31.89 31.72 64.93 65.10 31-94 3L79 64.95 Sulphur from the siphonate is high, doubtless from the trace of sulphur in potassium hydroxide. It will, therefore, be seen that it is possible to determine the metal from a pure zinc blende and that in the siphonate all the sulphur is completely oxidized and may readily be determined, 12 thus giving both chief constituents more easily and quickly than is possible by other methods. The ideal process of determining the sulphate ion electrolytically from this solution is yet to be found. The easy and accurate method of determining sulphate by titrating the acid liberated in the electrolysis of the salt is hardly applicable here in the large excess of potassium hydroxide re- quired. After weighing out the ore an hour should suffice for the zinc determination in a pure blende, but by carrying on a number of analyses at the same time it should be possible to considerably reduce the average time. APPLICATION OF THE METHOD TO OTHER ZINC ORES. The method having proved successful for a pure blende, its general applicability to zinc ores was now investigated. The nickel-plated platinum gauze served as cathode and the same vertical spiral of platinum as anode. The fusion was taken up in water and filtered through a Gooch crucible; before elec- trolyzing, this solution was heated. II. An ore from Phoenixville containing sphalerite and wurt- zite; less pure than I. Number Zinc blende Time heated Current Zinc found Zinc gram minutes Volts Amperes Time gram per cent. 39 0.5553 35 4-5 5 40 0.3230 58.17 40 0.5574 45 0.3249 58.li 41 0.7690 4 40 0.4606 59-89 42 0.4958 45 0.2978 60.06 43 0.5596 30 40 0.3357 59-99 44 0.4611 0.2776 60.20 45 0.5949 35 0.3558 59-8i Average 5 determinations 59-99 In Nos. 39 and 40 some zinc remained in the residue on the Gooch crucible, being held up by ferric hydroxide and other in- soluble substances. In Nos. 41-45, inclusive, the residue was dis- solved from off the filter in hydrochloric acid and reprecipitated with excess of potassium hydroxide, the second filtrate being added to the first. It is well to evaporate this hydrochloric acid solution nearly to dryness to drive off the excess of acid before reprecipitating with alkali. In these last five determinations, 13 which were successive, there was no zinc left in final residue on the Gooch, nor in the solution after electrolysis ; no iron in zinc deposit. A partial analysis made gravimetrically showed 3.65% gangue and 2.93% iron. It is evident then that if there is even so much as half of one per cent, of iron that it is necessary to dissolve and re- precipitate the ferric hydroxide to free it from zinc. From the beginning of the work it was thought that the separa- tion from the lead might be difficult, because of the formation of potassium plumbite, soluble in excess of alkali just as is the potassium zincate. After four or five non-concordant determ- inations were obtained on a mineral containing galenite, the ore was completely analyzed to find out just what were the condi- tions. III. An ore from Joplin containing sphalerite, wurtzite, gal- enite and calcite. On analysis this gave : Gangue 1.11% Sulphur S 21.05 (precipitated as BaSO*) Sulphur S 3.22 (from PbSO 4 ) Iron Fe Manganese Mn Zinc Zn Lead Pb Calcium carbonate CaCO 3 (from CaO) 100.12 From the many electrolytic determinations which were at- tempted the following may be given : Number Zinc ore Time heated Current Zinc found Zinc gram minutes Volts Amperes Time gram per cent. 46 0.3445 40 S 5 40 0.1503 43.62 47 0.4269 48 0.5409 49 0.4521 50 0.5882 51 0.5753 0.1838 43.05 0.2416 44.66 0.2157 47.71 5-5 35 0.2727 46.36 5 0.2592 45.05 In Nos. 46 and 47 the fusion was taken up in water and filtered directly from the crucible. The filtrate at first clear, clouded on dilution with wash water, the precipitate containing lead probably as the hydroxide. The residue on the filter was not dissolved and 14 reprecipitated and the zinc held back by the iron was compensated for by the lead deposited, the correct result being a mere coinci- dence due to the balancing of errors. In No. 48 the fusion was treated with water, poured into a beaker, diluted and boiled before filtering. This was done to save the additional filtration to remove the precipitate formed on dilution, but there was more lead in the deposit than in the two preceding cases. An increase in the original weight of the cathode after dissolving the zinc in the dilute sulphuric acid always indicated the presence of lead before a test had been made for it. An attempt was made in No. 49 to completely precipitate the lead as lead dioxide by adding potassium permanganate, precipi- tating the excess of manganate as the dioxide with alcohol, dis- solving and reprecipitating, but without success. In No. 50 a similar attempt at oxidation was made with hy- drogen peroxide and in No. 51 by fusion with addition of sodium peroxide with the same results. An effort was made to determ- ine the zinc by noting increase in cathode weight after treat- ment with sulphuric acid as lead, and deducting this from the total weight of the deposit, but the results were not reliable. It is noteworthy that the amount of lead deposited is always less than 5%, though the lead present in the ore is nearly 21%. This seems to be due chiefly to the formation of the white pre- cipitate mentioned above. There seemed to be no lead in the final zinc deposit when this procedure was adopted : The first zinc deposit containing some lead was dissolved in dilute sulphuric acid without drying and weighing. After adding alcohol to this solution it was filtered; to the filtrate an excess of potassium hydroxide was added and the solution warmed and electrolyzed. Time for fully working out this process was lacking. Further work will be done with these lead-zinc ores, and it is believed that the difficulty is one that may be overcome. * IV. More satisfactory results were obtained on the appli- cation of the method to a sample of franklinite from the New Jersey Zinc Co. 15 Number 52 53 54 It 57 Zinc ore gram 0.5173 0.5260 0.7822 0.5079 0.5826 0.5218 Time heated Current Zinc found Zinc minutes Volts Amperes Time gram per cent. 40 0.0804 15.54 0.0810 15.40 12 40 35 40 20 0.1233 0.0832 0.0938 0.0853 Average 4 determinations 16.15 On account of much manganese and iron hydroxides, the resi- due on the Gooch crucible was dissolved in hydrochloric acid and reprecipitated as in II. The addition of alcohol for the purpose of precipitating the hydrated dioxide from the maganate is neces- sary before both filtrations. In Nos. 52 and 53 there was trouble in the zinc deposition, the metal being deposited only around the bottom of the cathode and probably not entirely adherent. The hydrochloric acid solution had not been evaporated to remove excess of acid before addition of alkali and the potassium salts thus formed so greatly in- creased the conductivity as to render it difficult to obtain a high enough voltage. On evaporating off the excess of acid and re- plating the cathode with nickel, further trouble was avoided, though a higher amperage than that used with other ores was required. There was no zinc in the siphonate, nor iron in zinc deposit. No trace of maganese was found on either anode or cathode. As far as could be told with an excess of manganese and iron, there was no zinc in the residue on the filter. V. Two determinations on a sample of willemite from the New Jersey Zinc Co. show the method applicable to that mineral. Number Zinc ore Time heated Current Zinc found Zinc gram minutes Volts Amperes Time gram per cent. 58 0.5598 30 4.5 5 40 0.2808 50.16 59 0.6063 35 0.3018 49.78 The residue on the filter was dissolved and reprecipitated as usual. There was no zinc in final residue nor in siphonate ; no iron in zinc deposit. The samples from the New Jersey Zinc Co. were received through the courtesy of Mr. George C. Stone, to whom thanks are due. 16 The volume of the solutions electrolyzed varied from 100 c. c. to 250 c. c. This is not relatively of great importance ; in general, heat and concentration favor a rapid deposition. The fusions were done in nickel crucibles, which are but slightly attacked by the fused alkali, as many as fifty determinations being made in the same crucible. And this in spite of the fact that the fusions were usually heated longer than necessary, being sometimes al- lowed to run over a lecture hour. To eliminate platinum entirely from the apparatus required, an electrolysis of a standard zinc solution was made, using a spiral of nickel wire as anode and a cathode of nickel gauze. The de- termination was satisfactory and the anode being weighed both before and after the deposition showed no loss in weight. CONCLUSION. This work has brought out several points of interest: 1. That the metal in a pure zinc blende may be determined easily, rapidly and accurately in the electrolytic way after fusion with a caustic alkali, and that in the siphonate, or solution left after electrolysis, the sulphur is completely oxidized and ready for estimation in the usual manner with barium chloride. The method is shorter than others, whether electrolytic or not, when the metal alone is considered and possesses the additional advan- tage of the sulphur being available for determination. 2. The process is applicable to zinc ores generally with the possible exception of those containing lead. There is practically no opportunity for the entering of the personal element, for the current causes the deposit and the balance fixes its amount. Con- centration, temperature and exactly precise conditions have little influence and need not be constantly watched. The method works with equal ease for rich or poor ores. 3. By the use of nickel electrodes, not only more satisfactory deposits are obtained but the cost of apparatus is reduced to a minimum. This removes an objection frequently urged against electrolytic methods. The cylindrical cathode of nickel gauze dipping vertically into the solution in a beaker, is preferable to the dish or vertical sheet of metal. The possibility of the use of electrolytic methods in analyses of mineral products has been emphasized. Further work should be undertaken along this line. RETURN CIRCULATION DEPARTMENT 198 Main Stacks LOAN PERIOD 1 HOME USE 2 3 4 5 6 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS. Renewls and Recharges may be made 4 days prior to the due date. Books may be Renewed by calling 642-3405. DUE AS STAMPED BELOW I FORM NO. DD6 UNIVERSITY OF CALIFORNIA, BERKELEY BERKELEY, CA 94720-6000