Carbonated water exerts a corrosive action upon lead, therefore no syphon heads containing this contaminating metal should be permitted in use. Syphon heads have suffered and been measurably deteriorated by the presence of lead in the metallic composition from which they are made. Irresponsible and ignorant makers of this ingenious, handy article and invaluable assistant of the mineral-water manufacturer, are prone to push incomplete or pernicious goods upon the unsuspecting and unsophisticating carbonator. Of the danger lurking in the unrestricted employment of syphon heads susceptible of contamination when brought in contact with certain carbonated waters, experiment has furnished conclusive evidence. The use of syphons for lemonades, owing to the action of free tartaric acid upon lead, and the rapidity with which waters containing any free acid become charged with lead in syphons, must be condemned. In mineral water containing potash, drawn from a syphon, 0.0408 grain of lead per gallon was found to be present. Pure or plain carbonated water again drawn in a similar manner from the syphon gave 0.0816 grain of lead per gallon, or exactly double the amount found in the potash water, showing at once the well-known protective action that salts of the alkalies and alkaline earths have on lead.

These results, it may be truly said, are sufficiently high and alarming; still, when the water is drawn off in small quantities at a time, as is frequently the case, the results are found to be still higher. Thus, when potash water was so treated, 0.0455 grain of lead per gallon was found, while plain carbonated water, drawn off in small quantities, gave 0.0933 grain of lead per gallon, showing a very marked rise in both cases.

The cause of this increase in quantity of the lead appears to be owing, not so much to the lengthened period of contact between the liquid and the metal, as the fact that the nozzle of the syphon, being exposed to the atmosphere in a moist state, becomes rapidly oxydized, and is left in the most suitable condition for entering into solution, so that when merely small portions of the liquid are drawn off each time, a comparatively concentrated solution of lead is obtained. These results compare accurately with those obtained by examining the contents of a series of syphons of carbonated water for a physician, whose attention was drawn to the subject by detecting symptoms of lead poisoning in himself after he had been in the habit for some time of drinking such carbonated water.

There is no doubt of the evidence obtained of lead contamination in the use of syphon heads or tops, constructed of an amalgam, in which the pernicious metal was present in: undue proportion. Pure tin is comparatively soft and unfit for hard, wearing service, unless it is hardened by an alloy of some other metal. Lead and antimony are usually used to 26 effect this purpose, and when the proper amount is not exceeded no deleterious effects are traceable to their presence. An alloy of 99.90 pure tin submitted to severe tests, both with carbonic acid water under a heavy pressure and in contact with a concentrated solution of citric acid, showed no loss, sign of corrosion or trace of dissolved lead.

Block tin syphon heads, therefore, are the only ones that can be used without fear of contamination. The cost should be a secondary consideration in purchasing these goods. If the word of the manufacturer cannot be taken as a guarantee of the quality of his syphon tops, they may be easily tested for lead or other metals, by dissolving a section in hydrochloric acid, which yields a colorless solution. The presence of other metals in tin may be detected by treating the hydrochloric solution with nitric acid of a specific gravity of 1.160, first in the cold and afterwards with heat, until all the tin is thrown down in the state of insoluble stannic oxide. The decanted acid solution from pure tin leaves no residuum on evaporation. If, after the acid has been dissipated by heat, dilution with water occasions a heavy white precipitate, the sample contains bismuth; if, after dilution, a solution of sulphate of ammonium or of sodium produces a similar white precipitate, it contains lead; and if the clear liquid leaves a residuum, it contains copper. Other tests for lead in tin we have already given in Part III., when describing the properties of tin and its tests, to which we refer.

The popularity and consumption of mineral waters have been increased greatly where they are delivered in syphons, and the trade should in no wise hazard a diminution by countenancing any appliance which will not bear the closest scrutiny and severest tests as a container or dispenser of carbonated waters, either plain or saccharine.