Moore and Fitch (1912-13) considered the complement fixation test as employed by McFadyean and Stockman impractical. They stated that "The complexity of the method itself, the additional complication noted by Surface (inhibiting effect of excess of cow's serum) and by Thomson (apparent unreliability of inactivated serum), and the fact that a positive reaction does not give definite information as to whether the cow is infected and may abort or has been infected and already aborted, seem to render this method of diagnosis of little practical value."

Mohler and Traum (1914) made a thorough study of both tests, with the following conclusions: "In over 400 cases the agglutination test failed to show positive results in only one case where a positive reaction was ascertained by the complement fixation test, while, conversely, the latter test failed in four cases which were positive to agglutination, indicating in the latter tests that the agglutinins probably appear earlier in the infection than do the bacteriolytic bodies." They recommend the use of the agglutination test alone, due to its simplicity and the saving of time, except in doubtful cases, when the fixation test may be employed to advantage.

Rettger and White (1918) employed both reactions. They state: "From the beginning the results obtained by these two methods have been most gratifying." In a later paper (Rettger, White and Chapman, 1921) the importance of these tests is again emphasized. "The agglutination and complement fixation tests have been retained as the most satisfactory methods of following the course of infection in the herds . . . ; the real value of these tests has so impressed itself upon us during the whole course of the investigation that the entire work has been built around them as a supporting structure."

Seddon (1919) regards the agglutination test as thoroughly reliable. He believes that an agglutination titre of 0.01 is a priori evidence of infection with Bact. abortus. He calls attention to inhibition zones, and believes that at least six tubes of varying dilutions should be employed in each test.

Fitch, Boyd and Billings (1919) state that they "do not believe that the complement fixation test has any advantage over the agglutination test in the diagnosis of infectious abortion." They concluded that the blood reactions are of only limited value in controlling abortion and that it is safer to buy an animal with good breeding history than to rely on the results of the agglutination tests. Smith and his collaborators at the Rockefeller Institute have employed the agglutination test alone, believing that it follows the course of infection very closely.

Simms and Miller (1922) regard both methods as satisfactory, but advocate the use of the agglutination test because of its simplicity. Hart (1923), in discussing the agglutination test in particular, bares certain limitations. According to him, positive reactions mean past or present infection, and some cows which harbor the organism do not give positive tests.

Birch and Gillman in their recent work (1924) made monthly agglutination tests on all animals. They state that "The agglutination test coupled with individual or herd history gives information of great value to the purchaser. Consistent negative reactions and reactions only in very low dilution are strong evidence against active Bact. abortum infection . . . We would not accept the animals as a gift and add them to a clean herd if they were high reactors."

Boerner and Stubbs (1924) in their comparative studies of the agglutination and complement fixation tests found that "In over 95 per cent reactions obtained by one test were confirmed by the other." Further, "The reactions obtained by complement fixation were more distinctly positive or negative compared with the agglutination test"; and "A combination of the two tests is more accurate than either test alone."

Connaway (1921, 1923, 1924) regards the serological tests as specific for infectious abortion, and in his various publications places reliance on the tests. Likewise, Giltner, Cooledge and Huddleson and their co-workers have employed these methods and appear to have accepted them as specific and as valuable tests in the determination of Bact. abortus infection.

SEROLOGICAL METHODS EMPLOYED

The methods employed in the present daily routine are essentially the same as those described fully in Station Bulletin 93 (1918). Experience has shown that the older technique is capable of considerable refinement, however, and that it can be made thoroughly practical and dependable.

DRAWING OF BLOOD SAMPLES AND PREPARATION OF THE

BOVINE SERUM

Emphasis is again placed on the use of separate veterinary needles for each animal. The needles now in regular use in these laboratories are 16 gauge, length 1.5 inches. The needles are washed, then dried in an oven at a temperature sufficiently high to destroy all antibodies remaining from the blood.

The samples are cooled and immediately forwarded to the laboratory, where they are subjected to the serological tests. When, on rare occasions, some time intervenes between the bleeding and the testing, the serum is removed while fresh from the clot and preserved with 0.4% carbolic acid, or 0.2% tri-cresol. All samples are held in the refrigerator except during the time of transit.

The blood is drawn from the jugular vein, as is the common practice. The method of looping a rope around the neck has been abandoned because of its impracticability. The jugular vein can be brought into prominence easily by pressure of the thumb or fingers.

THE MACROSCOPIC AGGLUTINATION TEST

The agglutination antigen is prepared from 48 hour slant agar* cultures of at least three different strains of Bacterium abortus of known agglutinability. The surface growths are washed from the agar with physiological saline solution (0.85%) containing 0.3% phenol. After filtration through cotton the antigen suspensions are further diluted with saline solution until they match tube 1.0 of the McFarland nephelometer scale (1907).

The agglutination tests are made in 3" x 2" test tubes. To 3cc amounts of the antigen suspension serum is added to make final dilutions of 1:50 and 1:100, namely, 0.06cc and 0.03cc of the undiluted serum. The corked and thoroughly shaken tubes are incubated at 37°C. for 48 hours. Following the first recording of results the tubes are allowed to stand at ordinary room temperature for another 24 hours, at the end of which time a second and final reading is made.

In the routine work the results are recorded as positive (+), doubtful (?) or negative (o). In experiments in which agglutination titres are sought the four plus system of recording is employed. Tubes are considered positive only when both clearing and flocculation are complete, and negative when there is no indication whatever of either change.

Shorter periods of incubation have proven less satisfactory. Incubation at 37°C., rather than some high temperature, has been found desirable, owing to the more general availability of 37°C. incubators.

Experience may prove that somewhat higher dilutions should be employed than have been used here, though remarkably close agreements have all along been obtained between results of the agglutination and the complement fixation tests. It should be said, however, that occasionally non-specific doubtful reactions are obtained in the lower (rarely in both) of these dilutions in animals that are subsequently found to be free from Bact. abortus infection.

COMPLEMENT FIXATION

The technique of the complement fixation test appears to the beginner exceedingly complicated, but it should become comparatively simple with practice. The fixation method for infectious. abortion has long passed the experimental stage; however, it is only with the newer improvements of technique and with constant attention to the finer details that it can be made thoroughly successful.

Fairchild peptone has been employed during the past five years in all agar used for the production of Bact. abortus antigen in the agglutination and fixation work.

Such cultures should be prepared with inoculum from 24 hour preliminary cultures.

The production of a satisfactory antigen is without doubt the most difficult and uncertain phase of fixation work, but even this should no longer constitute any drawback.

Sheep's Corpuscles

Sheep's blood is drawn and defibrinated in the usual manner. It is filtered immediately through absorbent cotton. All this should be done under as nearly aseptic conditions as possible. The filtered blood is kept in the ice box. For immediate use a portion is treated in the following manner. The corpuscles are removed by precipitation in a high speed centrifuge (3,000 to 4,000 revolutions per minute), and washed at least three times with physiological saline solution. Unless the corpuscles are to be kept longer than four or five days, they need not be preserved. It is, in fact, better to procure fresh corpuscles, when needed, than to depend upon any known method of chemical preservation. (See Wenner, 1918.)

Hemolysin

Very strong hemolysin may be obtained by the following method. Two strong rabbits are given intravenous injections of undiluted washed sheep's corpuscles according to the following schedule:

Eight days after the last injection a trial bleeding is made from the marginal ear vein. If the serum is found sufficiently potent the rabbits are bled to death or enough blood is taken from the ear vein as is desired for stock hemolysin. The latter method should yield all the serum needed, at least if the bleeding is repeated on two or three successive days, and if both ears are used.

The serum is allowed to separate from the clot, pipetted off, and treated with 0.4% phenol. The potency will be retained for many months, when stored in the refrigerator. Titrations should be made, however, at intervals not exceeding three or four months. The hemolytic titer (unit) should be at least 0.05 in 1:50 dilution.

Suspensions of Corpuscles

For all purposes of titration and for the complement fixation test 2% suspensions of red corpuscles are prepared by adding 2cc of fresh washed erythrocytes to 98cc of sterile saline solution.

Titration of Hemolysin

Rabbits' serum which is less than five days old should be inactivated by heating at 56°C. for 20 to 30 minutes. This suffices to remove all complement that may be present.

Potent serum should require a dilution of at least 1:50 with saline solution to furnish the proper range for titration and fixation work. The amounts of reagents and steps involved in hemolysin titration are given in the following sample table. Each tube is well shaken and incubated for 30 minutes in a water bath held at 37°C. The smallest amount of serum required to bring about complete laking constitutes the hemolytic titre as unit.

The complement has a titer in the above instance of 0.03. In order to secure complete hemolysis 1.5 units are employed. A slight loss of complement from absorption and other neutralyzing factors may occur, hence a small allowance is made. The hemolytic titre is 0.05.

Complement

The Wenner method (1918) of bleeding guinea-pigs has been used throughout the investigations. It has proven itself to be most satisfactory, and the means of an enormous saving in guineapigs. The technique is comparatively simple and may be acquired easily.

Large male guinea-pigs are usually employed. They are etherized, just enough anaesthetic being given to render the animals insensible to pain. The hair on the neck is clipped closely, the clipped portion washed with alcohol and rubbed almost dry with absorbent cotton. The guinea-pigs are then suspended from a tall support by a cord drawn around one of the hind legs. The head is grasped with the left hand, the thumb and forefinger being used to draw the skin tightly around the neck. The animal is turned, ventral side up and head pushed back and up. A short transverse incision 10-15 mm long is made with sharp surgical scissors midway between the lower point of the jaw bone and the trachea.