THE MICROSCOPICAL EXAMINATION OF WATER.

In the special report of the Board upon the investigations conducted at the Lawrence Experiment Station on the purification of sewage and water, Professor Sedgwick, Biologist of the Board, has given an historical account of the various methods hitherto employed for the microscopical examination of potable water, and has described an entirely new method introduced by himself for the work of the Board, and afterwards much improved by George W. Rafter, C. E., of Rochester, New York.*

This method, which is now known as the Sedgwick-Rafter method, has been constantly employed in the regular examinations of the drinking waters of Massachusetts for the State Board of Health since Nov. 6, 1890. By its aid and under the supervision of Professor Sedgwick I have made more than 3,000 examinations of potable waters, my whole time having been given to this branch of the biological work.

In the present paper I propose, first, to describe the method in detail for it is still comparatively unknown-and to suggest certain improvements; and, second, to consider at some length the possible sources of error which its use may involve. I desire to acknowledge my obligations throughout to Professor Sedgwick.

It should be clearly understood at the outset that the microscopical examination of water has nothing to do with the smallest forms of aquatic life (the bacteria) or with the largest forms (such as fishes, frogs, pond-lilies, etc.). For detecting and counting the bacteria special methods, known as "cultures," are indispensable. For recognizing the larger forms the naked eye is sufficient. Between

W. T. Sedgwick. Recent Progress in Biological Water Analysis. Journal of the New England Water Works Association, September, 1889.

George W. Rafter, C E. The Biological Examination of Potable Water. Proceedings, Rochester Academy of Sciences, 1890. Rochester, New York.

W. T. Sedgwick. Report of the Biological Work of the Lawrence Experiment Station. Special Report of the Massachusetts State Board of Health on the Purification of Sewage and Water, 1890.

these limits, however, there is a vast host of minute plants and animals, nearly or quite invisible to the naked eye, yet readily studied (as bacteria are not) by the aid of microscopes of ordinary power. These forms abound in lakes and ponds, in open reservoirs and in rivers.

To distinguish them from the bacterial organisms, which cannot so satisfactorily be studied by the aid of the microscope alone, Professor Sedgwick has proposed that they be called the Microscopical Organisms, the whole group of small or micro-organisms being sub-divided and defined as follows:

By this simple and natural classification all of the lowest plants and animals are grouped together in one great class, viz.: the Micro-Organisms. Most of these, though very small, are still easily studied by the aid of the compound microscope. The members of one group only, the bacteria, are so extremely minute that the microscope fails to reveal most of their peculiarities. The microscopical organisms, therefore, include all of the lowest animals and, excepting the bacteria, all of the lowest plants; but until the Sedgwick-Rafter method was devised there was no satisfactory method for the collection, enumeration and comparison of these organisms in different specimens of water. This method, originally devised to meet the exigencies of the work of the Board, though by no means absolutely perfect, has been found to work very well.

DETAILED DESCRIPTION OF THE SEDGWICK-RAFTER METHOD FOR THE MICROSCOPICAL EXAMINATION OF DRINKING WATERS. The method consists, first, in the concentration of all the organisms (except the bacteria) in a measured quantity of the water to be examined by its filtration through sand upon which the organisms are detained; second, in the separation of the organisms from the

* See Report of Prof. Sedgwick cited above, page 797.

sand by agitation in a small and known volume of distilled water, followed by rapid decantation; third, in the microscopical examination of a measured part of the decanted fluid in which the organisms are suspended; and, finally, from the results of this examination, an estimation of the total number and kinds of organisms present.

In the routine work of the Board the measured quantity taken for examination is usually 500 cubic centimeters. This amount is withdrawn before any water is taken from the bottle for other purposes from the large four-liter bottles in which samples of water from the several public water supplies of the State regularly arrive for analysis. Before withdrawing the 500 cubic centimeters, moreover, the bottle is thoroughly agitated in order to dislodge any sediment from the bottom, and to obtain a uniform mixture of the organisms and water.

The microscopical examination is then conducted as follows: The tube of an ordinary six-inch glass funnel is plugged by a perforated rubber stopper, over the upper (smaller) end of which has been laid a circular piece of fine silk bolting-cloth, cut out by a wad-cutter. The stopper should flare but little, and its smaller end should tightly fit the funnel stem, leaving no space between the glass and the rubber. Over the hole in the stopper lies the bolting cloth, which, if laid upon the stopper and moistened, before the latter is inserted into the funnel tube, will adhere to it and form an excellent sieve-like support for the sand afterwards to be poured in from above. Quartz sand, sharp, dry and thoroughly clean, is used for the filtering material, and enough is taken at each time to fill the tube of the funnel to a depth of about one-half an inch above the plug. The sand should be fine enough to pass through a brass seive having 60 meshes to an inch, but not through one having 120 meshes to an inch. It is washed into position and saturated with a small amount of distilled water from a wash-bottle, special care being taken that the distilled water used is fresh and free from organisms. By this previous saturation air-bubbles in the sand are avoided. The funnel thus stoppered with a perforated plug and charged with sand constitutes the filtering apparatus for collecting and concentrating the organisms.

The water to be examined is now thoroughly shaken and poured into the funnel. It runs through the sand more rapidly at first than later, when the filter becomes more efficient through clogging of the inter

stices between the surface grains. To avoid a small loss of organisms the first 50 cubic centimeters of the filtrate may be poured back into the funnel and made to pass through the filter a second time, although in actual practice this refinement is usually omitted because the error is very small.

After the water has all passed through, the plug is carefully removed and the sand, with whatever organisms have been detained, is washed down into a test tube by five cubic centimeters of distilled water delivered from a graduated pipette. By a rotary motion the contents of the test tube are thoroughly agitated and the organisms are thus freed from the sand. On bringing the tube to rest, the sand, being heavier, quickly settles leaving most of the organisms for an instant in suspension. The supernatant fluid is now rapidly decanted into a second test tube, and, if the process is skillfully done, carries with it most of the organisms, but leaves the sand behind. Care being taken to secure an average sample from about midway between the top and bottom of the five cubic centimeters, one cubic centimeter of the decanted fluid is next withdrawn by a pipette and transferred to the counting plate. This consists of an ordinary glass slide upon which has been cemented a rectangular brass border (20 x 50 millimeters), enclosing an area of exactly one thousand square millimeters. This open chamber is next covered by a piece of thin glass, and, if the brass border is exactly one millimeter in height, must contain exactly one cubic centimeter. As the brass border is cemented to the slide the chamber usually contains slightly more than this and a small bubble of air may, after covering, sometimes remain in the chamber; but if thin cement has been used and if all precautions be taken, the chamber may be made to contain almost exactly one cubic centimeter.

Great care must be taken from the very beginning to secure average samples and at the end to obtain an even distribution of the organisms upon the counting plate. I have found that the latter may be obtained by proper arrangement of the cover glass. The glass is laid diagonally over the cell in such a manner that an opening about four millimeters wide is left in a corner at one end. At the corner opposite there is a small opening for the escape of air. The sample to be examined issues from the pipette into the larger opening and by capillary action completely fills the cell, and gives an even distribution of the organisms contained in it.

The counting chamber, having been filled with the sample to be examined, must obviously contain one cubic centimeter of the decanted fluid, in which there must be one-fifth of the total number of organisms separated by filtration from the original sample; so if 500 cubic centimeters, for example, were originally taken, the contents of 100 cubic centimeters are now ready for inspection within the counting chamber. We may, therefore, at our convenience, examine qualitatively the organisms present, using a comparatively low power, e. g., the B or C objective of Zeiss.

To make the examination quantitative it is only necessary to place in the ocular of the microscope a unit measure of the area of the counting plate. This is readily effected by placing upon the diaphragm of the eye-piece an ocular micrometer consisting of a glass disc upon which is marked a square of such size that it can be. adapted, with the objective used, to cover exactly one square millimeter of the counting plate. In order to exclude all objects not falling within the square millimeter to be examined, the disc, if it is preferred, may be blackened outside the ruled square. The square alone answers very well, and was used for a long time; afterwards it was subdivided by two fine lines into four smaller squares, and more recently it has been still further subdivided with advantage. It is hardly necessary to add that this ocular square must be carefully standardized for different powers. Having thus standard. ized it, it is only necessary to remember that the field or aperture of the ocular diaphragm corresponds to an area of one square millimeter and also covers a volume of one cubic millimeter upon the counting plate.

If the contents of this cubic millimeter be noted, and other portions of the plate be similarly examined, we are obviously able to arrive at an approximately accurate idea of the kinds and numbers of microscopical organisms before us; but, inasmuch as the plate contains 1,000 cubic millimeters, it is clearly impracticable to count, or even to scrutinize, the contents of all of these. In practice, only 20" fields," or cubic millimeters, are ordinarily examined, i. e., one-fiftieth of the whole number. This means that if 500 cubic centimeters were originally taken, and one-fifth of all the organisms in the original measured quantity is before us on the counting plate, evenly distributed throughout 1,000 cubic millimeters, we might, if time allowed, actually examine the contents of every one of these.