UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA PULLORUM DISEASE (Bacillary White Diarrhea of Chickens) J. R. BEACH and S. T. MICHAEL BULLETIN 486 January, 1930 UNIVERSITY OF CALIFORNIA PRINTING OFFICE BERKELEY, CALIFORNIA 1930 Digitized by the Internet Archive in 2012 with funding from University of California, Davis Libraries http://www.archive.org/details/pullorumdiseaseb486beac PULLORUM DISEASE (BACILLARY WHITE DIARRHEA OF CHICKENS) J. K. BEACHi and S. T. MICHAEL* INTRODUCTION The term 'bacillary white diarrhea' is neither sufficiently descrip- tive nor sufficiently inclusive to apply to infections of both chicks and adults with Bacterium pullorwm. As a substitute for 'bacillary white diarrhea' the term 'pullorum disease/ as applying to all infections with Bad. pullorum, has been suggested by the Pennsylvania State Department of Agriculture (1) 3 and was endorsed by Rettger (2) in a recent publication, and also by a "conference of laboratory workers in pullorum disease (B. W. D.) eradication " (3) held in May, 1929, at Yale University. This designation seems very appropriate and is used hereafter in this discussion. The increasing recognition by poultrymen of the economic impor- tance of pullorum disease, the growing traffic in day-old chicks which makes possible wide-scale dissemination of a disease of this character, and the existence of decided differences in the opinions of investigators regarding the efficacy of the control measures extensively practiced have caused this disease in recent years to receive much attention from investigators and to be given much prominence in veterinary and poul- try meetings and publications. In this discussion an attempt is made to give an impartial presentation of the various phases of pullorum disease based upon the results of studies conducted at this and other research institutions. CAUSE AND NATURE OF THE DISEASE Pullorum disease is caused by infection with Bacterium pullorum and affects both chicks and adults. Chicks have an acute form of the disease from which many succumb after a brief illness. The term 'white diarrhea' or 'baeillary white diarrhea' by which the disease has formerly been known suggests that diarrhea is a particularly charac- teristic symptom. However, diarrhea is often absent in pullorum disease and is frequently present in other common diseases of chicks. In adults the infection is usually localized in the ovaries and rarely produces visible symptoms. i Associate Professor of Veterinary Science and Veterinarian in the Exper- iment Station. 2 Kesearch Assistant in Veterinary Science. 3 Superscript numbers in parenthesis refer to "Literature Cited,' ' pages 28-31. University of California — Experiment Station DESCRIPTION AND DIAGNOSIS OF THE DISEASE IN CHICKS Symptoms. — Diseased chicks do not exhibit any symptoms that are of particular diagnostic value. A small amount of chalk-like material may be seen adhering to the fluff below the vent of many of the chicks, and occasionally such collections around the vent attain considerable size. This is the probable explanation of the name 'bacillary white diar- rhea'. Even this suggestion of diarrhea is very often absent, however, and is perhaps just as frequently seen accompanying conditions other than the one under discussion. A high mortality in a flock of cor- rectly brooded chicks during the first two weeks of their lives strongly suggests pullorum disease but is not a definite indication of it since there are other causes of excessive mortality among young chicks. According to the recent observations of Bunyea and Hall, (4) an abrupt rise in the number of deaths six to eight days after hatching, followed by a considerably decreased yet abnormally high mortality up to two weeks of age is more suggestive of pullorum disease than a high mortality during the first few days after hatching. Post-Mortem Findings. — In nearly all of the earlier descriptions of the post-mortem appearance of dead chicks, livers uniformly yellow or ochre colored or yellow with bands or streaks of normal color and unabsorbed yolks are given as characteristic lesions. It is now rather generally recognized, however, that such lesions are commonly seen in chicks that die during the first week after hatching regardless of the cause of death; that unabsorbed yolk is usually present in normal chicks of that age; and that neither unabsorbed yolk nor yellow livers are often found in chicks that die after the first week. In chicks that die from pullorum disease during the second week, more characteristic changes in the organs are apt to be present. These consist of gray foci in the liver, gray nodules in the heart wall, and gray or yellowish nodules in the lungs. Both Doyle and Mathews (5) and Bunyea and Hall, (4) who have furnished the most recent contri- butions on the pathology of pullorum disease, consider these as charac- teristic of the disease. Doyle and Mathews (5) consider any type of pneumonia in chicks to be strongly suggestive of pullorum disease, but in this opinion Bunyea and Hall (4) do not concur. The intestines do not commonly show characteristic changes, although Bunyea and Hall (4) observed in a few chicks a chronic in- testinal type of infection with necrotic foci in the large intestine. BUL. 486] PULLORUM DISEASE 5 Stafseth and Johnson (6) have made an interesting- report of atypical symptoms and lesions from Bacterium pullorum infection in one flock of chicks. In this case the chicks developed a gelatinous exudate located subcutaneously and in the body cavities. Diagnosis. — The studies of Doyle and Mathews (5) led them to conclude that pullorum disease of chicks could usually be diagnosed solely by gross lesions. It has been the observation of the writers, however, and also of others who have had extensive experience with the disease, including Bunyea and Hall, (4) that distinctive gross lesions are so frequently absent in acute cases of bacillary white diarrhea that a bacteriological examination is usually necessary for a definite diagnosis. DESCRIPTION AND DIAGNOSIS OF THE DISEASE IN ADULTS Bacterium pullorum infection among flocks of adults is very wide- spread. To illustrate this fact, the results of some agglutination tests made at this station in 1925 might be cited. During that year, 108 flocks, containing 36,000 birds, were tested for the first time. Some reactors were found in all except four flocks. In view of the fact that there had been very little loss from pullorum disease among chicks in this section, the results were somewhat surprising. The Acute Type of the Disease. — Occasional outbreaks among adults of an acute septicemic type of infection with Bacterium pullo- rum have been reported. It is probable that some of the earlier reported outbreaks of this nature were in reality due to some other but closely related species of organisms, which, because of inadequate bacteriological procedures, were not differentiated from Bad. pullo- rum. There is little doubt, however, that Bact. pullorum is frequently the cause of fatal septicemia in individual birds although probably severe losses from this cause within a short time are relatively uncommon. Bunyea (T) has reported the occurrence of outbreaks of inflamma- tion of the oviduct among pullets as a result of Bacterium pullorum infections, manifested by eggs with an offensive odor, blood-spotted and blood-smeared eggs, soft-shelled eggs, a profuse white discharge from the vent, and the practice of cannibalism. Occurences of this nature, however, cannot be ascribed to Bact. pullorum infection from symptoms alone. The writers have made numerous examinations of similar cases without being able to incriminate any organism as the causative factor. 6 University of California — Experiment Station The Chronic or Localized Type of the Disease. — Bacterium pullo- rum infection of adults usually becomes localized in some organ where it remains dormant and produces no external evidence of disease. In males found to be harboring the organism, it has been isolated from Fig. 1. — Ovaries of a hen infected with pulloram disease. Numbers 1 and 2 arc normal yolks. Numbers 3 to 9 and many of the smaller yolks are abnormal because of the infection. the testes, spleen, gall bladder, pericardial sac, and heart muscle. In the writers' experience, it has been isolated more frequently from the pericardial sac, where it has produced an exudative pericarditis, than from any other source. In females, the ovary is the organ most fre- quently involved. The prominent changes here are abnormal yolks BUL. 486] PULLORUM DISEASE 7 (fig. 1). These abnormalities include bloody and caseated ovules; cysts attached to the ovary; small, partially solidified, blood-tinged yolks; yolks of various sizes with capsules but partially filled with thick yellow or greenish-yellow liquid; solidified, angular, yellow or greenish-yellow or blood-tinged yolks ; and yolks with a thick opaque capsule containing yellow, semi-solid, oily material or a clear yellow, oily liquid with white flakes in suspension. Localization of the infection may also occur in other and some- times very obscure places, either in addition to or in the absence of ovarian lesions. Isolation of the organism from abdominal cysts not attached to the ovary or oviduct and from the pericardial sac is relatively common. There has been one case recently reported (8) in which it was isolated from the middle ear only, with no evidence of infection in any other part. As with chicks, the presence or absence of characteristic lesions does not definitely indicate the presence or absence of infection, and bacteriological examination must usually be resorted to in establishing a diagnosis. A fairly accurate means of detecting pullorum disease in the living fowl is afforded by the agglutination test. A full discussion of the test is given later. The Effect of the Disease on the Production and Hatchability of Eggs. — Bacterium pidlorum infection of adults, although usually localized and not responsible for excessive loss from death of birds, causes loss in other ways which is of serious consequence. This is the effect of the infection on production and reproduction. It has been rather definitely established that infected fowls, as a rule, lay fewer eggs than non-infected fowls of like quality. The most serious consequence, however, is that eggs laid by infected hens may contain the infection. As a result of such infection in eggs the fertility of eggs is decreased, many embryos die in the shell during incubation, and many chicks are infected when they hatch. A good example of the effect of the infection on hatchability of eggs recently came to the attention of the writers. In this case, a flock of 298 hens, all with a trapnest record of more than 250 eggs during their first laying year, were divided into two groups, one of 202 non-reacting and the other of 96 reacting birds, on the basis of their reaction to the agglutination test. It is probable that all of the reacting birds were infected but since this was the first time they had been tested and the percentage of reactors was high, there were undoubtedly infected birds remaining in the non-reacting group. From December 7, 1928 to February 1, 1929, 2159 eggs from the non-reacting group and 463 eggs from the reacting group were set. The results are given in table 1. 8 University of California — Experiment Station TABLE 1 Tabulated Kesults of Incubating 2,159 Eggs from Fowls that did not Keact and 463 Eggs from Fowls that did Keact to the Agglutination Test Results of incubal ion Per cent b Group Per cent of eggs infertile Per cent a dead germs Per cent a dead in shell Per cent a cull chicks Per cent 3, vigorous chicks mortality during first two weeks Non-reactors Reactors 14.7 14.7 4.0 7.1 22.1 36 4 3.1 10.6 69.1 45.6 0.8, 11.7* a Percentage of fertile eggs. b Percentage of vigorous chicks. It can be seen from this data that the loss to the poultry breeder from the decrease in hatehability of eggs resulting from pullorum disease in hens may reach serious proportions. The question of the infection in chicks hatched from eggs laid by infected hens will be considered later. SOURCES OF THE INFECTION IN CHICKS The origin of the infection responsible for most outbreaks among flocks of chicks lies in chicks hatched from infected eggs laid by infected hens. It is probable that, as a rule, there are only a few infected birds among a batch of chicks at the time of hatching. The infection can rapidly spread, however, from the few to many others before the chicks leave the incubator or after they are placed in shipping boxes or in the brooder. It has been shown, particularly by Hinshaw, Scott, and Payne, (9) that with the types of incubators in which there is a forced circulation of air the infection may be widely disseminated by means of down from the infected chicks. With the ordinary type of incubator, consisting of a series of small com- partments, the spread of infection is probably ordinarily confined to the compartment in which the infected chicks are located. Incu- bator-acquired infection may enter the bodies of the chicks by way of either the digestive or the respiratory tracts. Infection by the latter route is probably responsible for the occurrence of lesions in the lungs. Other sources of infection in chicks are contaminated incu- bators, shipping boxes, brooder houses, or brooding equipment. Such sources can be readily eliminated by sanitary measures. Variation in the Extent of the Infection Among Progeny of In- fected Flocks. — Much variation has been observed in the occurrence of pullorum disease among different hatches of chicks from flocks in BUL. 486] PULLORUM DISEASE 9 which there were presumably infected hens and even among different lots of chicks of the same hatch that were shipped to different poultry- men. Many breeders and hatcherymen and even some investigators have, therefore, questioned whether the infected hen was really dan- gerous as a source of infection among chicks. Careful consideration of some of the known facts regarding the infection in hens and its manner of dissemination would seem to afford a reasonable explana- tion for variations of this nature. These are as follows : 1. A wide variation has been shown to exist in the frequency of the occurrence of infection in eggs laid by individual infected hens. 2. Many infected eggs fail to hatch because of infertility or death of the embryo in the shell. 3. Many of the chicks that hatch from infected eggs die or are destroyed soon after they emerge from the shell and before there has been much opportunity for dissemination of infection among the healthy. Tt is, therefore, conceivable that even though a flock may contain numerous infected hens, very few infected eggs may be produced or hatch and loss from pullorum disease may occur inf requently or irreg- ularly among the offspring from it. It must be borne in mind, how- ever, that exactly the reverse may occur. Numerous infected eggs may be laid from which infected chicks may be hatched and conditions may be such that the infection will be rapidly and widely dissemi- nated among the others of the hatch. SOURCES OF THE INFECTION IN ADULTS Many of the survivors of outbreaks of pullorum disease among chicks continue to harbor the causative organisms, which become localized in some organ as previously described. This is considered the most common source of infection in adult flocks. Fowls can, however, and frequently do acquire the infection and become carriers after they have matured. The practice of feeding uncooked infertile and dead germ eggs from incubators has produced extensive infection in some flocks. It has also been demonstrated that healthy hens may acquire in- fection from association with infected hens. As one phase of an experiment conducted by the writers, 37 non-reacting hens were kept in a pen with 61 reacting hens for nineteen months, during which time agglutination tests were made at monthly intervals. These birds had 10 University of California — Experiment Station been tested once each month during the preceding year. The 37 non-reacting hens were housed separately during this time and their reactions had been uniformly negative. Twelve of these became reactors to the agglutination tests after they were placed with the group of reactors. All were later destroyed. Bacterium pullorum was isolated from nine of the twelve. Failure to isolate the organism from the other three birds was probably due to the inadequacy of our efforts rather than to absence of infection. The interval between the time of placing the two groups together and the first reaction of the birds that became infected varied from three to nineteen months. Brunett (10) conducted a similar experiment in which he kept 17 non-reacting with 20 reacting hens for seven months without any transmission of the infection. Different results might have been obtained if the experiment had been continued for a longer period. To what extent males may be responsible for the spread of infec- tion among hens is not yet clearly understood. The actual trans- mission of the disease to hens by infected males has not, in so far as the writers' knowledge extends, been demonstrated. Nevertheless, an infected male must be regarded as a potential source of infection of hens with which he may be associated. There is considerable evidence that the infection will spread more rapidly among hens when males are present in the flock than when only females are present. The role of the male in this case is probably that of a mechanical carrier of infection from hen to hen. CONTROL OF OUTBREAKS OF THE DISEASE IN CHICKS There is little evidence to indicate that efforts to reduce losses from outbreaks of pullorum disease among chicks have, as a rule, been highly successful. Many of the seemingly good results from methods that have been tried are probably due to the facts that the mortality in outbreaks varies between wide limits and many times is low, and that the deaths usually have become greatly decreased or have entirely ceased by the time the chicks are from two to three weeks old. The employment of sanitary measures are undoubtedly of benefit if properly carried out. These would consist of the usual cleaning and disinfection of the houses and equipment, the prompt removal and destruction of sick or dead birds, and dividing the chicks into as small units as possible. This last -mentioned measure might be highly effective in preventing heavy losses if it were done early in the BUD. 486] PULLORUM DISEASE 11 outbreak, in case there were not many infected birds remaining in the flock after the visibly sick ones had been removed and these were contained in one or two of the small units. In controlled experiments with artificially infected chicks, drugs and chemicals have been found of no benefit when taken into the alimentary tract and, therefore, ineffective in reducing mortality among chicks that were actually infected. Some chemicals, however, particularly the hypochlorite solutions, are of value as disinfectants in drinking water to prevent the spread of infection through this route. Such are the conclusions of May and his associates 01 - 12) after a series of experiments concerning the effect of chemicals in the control of poultry diseases in which potassium permanganate, hydro- chloric acid, mercuric chlorid, sulfocarbolates, sulfuric acid, resorcin, and several hypochlorite solutions were tried. The hypochlorite solu- tions were used in sufficient amount to provide from 0.020 to 0.035 per cent available chlorin in the drinking water. PREVENTION OF THE DISEASE IN CHICKS The prevention of pullorum disease, like that of other infectious diseases, lies in removing the sources of infection. These are, as previously stated, contaminated incubators, shipping boxes, and brooders, infected chicks, particularly before removal from the incu- bator, and carrier hens. The incubators, boxes, and brooders can readily be eliminated as sources of infection by thorough clean- ing and disinfection before each time they are used. The dis- covery that infection was disseminated in incubators by the down of chicks hatched from infected eggs at first appeared to present a difficult problem. Investigations conducted separately by several workers (Gwatkin, (13) Coon, (14) Payne, (15) and Dakan and Speer (16) ), however, appear to have provided a solution. This consists in liberat- ing formaldehyde gas 4 within the incubators at suitable intervals during the incubation period and while hatching is taking place. It is said that this procedure, if correctly done, will not be injurious either to the hatchability of the eggs or to the chicks and will prevent dissemination of the disease through the incubator by destroying the infection on the down of chicks that have hatched from infected eggs. This treatment cannot, of course, be expected to destroy the organisms within the bodies of the infected chicks and, therefore, they remain 4 The detailed recommendations for disinfecting incubators with formaldehyde gas will be furnished on request by the Division of Veterinary Science, University of California, Berkeley, California. 12 University of California — Experiment Station sources of infection after they are taken from the incubator. Careful inspection of the chicks for the purpose of detecting and removing" those that are weak will serve to remove many of the infected ones and will accordingly reduce the likelihood of serious losses from pullorum disease in the brooder. THE TUBE OR SLOW SERUM AGGLUTINATION TEST FOR THE DETECTION OF MATURE FOWLS THAT ARE CARRIERS OF BACTERIUM PULLORUM There is now left for consideration the means of eliminating; the most serious source of pullorum disease in chicks, i.e., the carrier hen. In 1913, Jones (17) determined that the agglutination test was of value for this purpose. His findings were confirmed by others and 'B. W. D. testing' has been extensively practiced for several years. Fig. 2. — Method of drawing a blood sample for the agglutination test. Description of the Test. — For the readers who are not familiar with the procedure of making an agglutination test, the following brief description is given : A sample of blood is drawn from each bird in small vials or test tubes (fig. 2). After the blood has clotted, the vials are put in a cool place until the serum or clear liquid portion of the blood has separated from the clot. A minute quantity (usually from 0.01 to 0.04 cubic centimeters) of the serum of each sample is BUL. 486] PULLORUM DISEASE 13 mixed with a small amount (usually 1 cc) of test fluid or antigen in a test tube. The antigen consists of cultures of Bacterium pullorum in suspension in 0.85 per cent of salt solution containing a small amount of phenol or other preservative and is slightly turbid. The serum-antigen mixture is allowed to stand for twenty-four hours or longer at incubator or room temperature before a reading of the results of the test is made. If the blood serum contains what is known as agglutinins, the organisms in suspension in the antigen are caused Fig. 3. — A positive (left) and negative (right) reaction to the tube or slow agglutination test. to clump together and sink to the bottom of the tube with consequent clearing of the fluid. This is known as a positive reaction (fig. 3). If the blood serum contains no agglutinins, the antigen remains turbid. This is known as a negative reaction (fig. 3). The blood serum of many of the fowls with pullorum disease contains agglutinins and gives a positive reaction to the agglutination test. Discrepancies have existed in the results obtained from the test in different parts of the country. As a result, there is considerable variance in opinions held regarding its efficacy, some workers even expressing the belief that its inaccuracies are so great that it is of little or no practical value. Studies that have been made in recent years are doing much to relieve this situation by bringing about a better understanding of the virtues and limitations of the test and improvements in the methods of making it. 14 University of California — Experiment Station Cloudy Reactions. — For several years, the occurrence of excessive turbidity (cloudy reactions) in many tubes, which made accurate inter- pretation of the results of the test in such tubes practically impossible, was an important source of error. It is now known that this is due to the precipitation of some substance (fat or protein) from the blood serum by the phenol preservative in the antigen, and means of avoid- ing it have been found. These include the reduction of the amount of phenol in the antigen, the substitution of formalin for phenol in the antigen, and the addition of sodium hydroxid to the antigen. The last named, which was developed by Mathews (18) and Mailman, (19) appears to be the most satisfactory. It is said not only to eliminate excessive cloudiness, but also to make the antigen slightly more sensitive. Variation in Methods. — Variation in methods of performing the test in different laboratories has undoubtedly been responsible for many of the discrepancies in results. A few of the variations in methods are as follows: Some laboratories use only one dilution of serum and antigen, others use two or more dilutions ; in most labora- tories the tests are kept at incubator temperature for at least twenty- four hours, but in some this is not considered necessary; the time of making the final reading varies from twenty-four to seventy-two hours after the test is started; the density of antigen varies to a considerable degree in different laboratories. The Poultry Disease Committee of the United States Live Stock Sanitary Association (20) has attempted to effect some standardization of methods by making tentative recommendations for carrying out all of the procedures necessary for the performance of tests. The com- mittee further recommended that flocks be tested at a number of laboratories in accordance with the proposed tentative standards, so that definite information might be obtained regarding the practical value of the proposed technique. To what extent this request has been complied with is not known. Necessity for Repeated Tests to Eradicate the Disease from a Flock. — It is well known that repeated tests are necessary for the elimination of all of the fowls in a breeding flock that harbor Bacterium pullorum. For many years, the explanation given for the failure of a single test to detect all of the infected fowls was either that certain of the birds had acquired the infection too recently for the production of sufficient agglutinins in their blood to cause an agglutination reaction or that certain of the birds had become infected after the test, either from association with infected birds or from contaminated litter and soil. BUL. 486] PULLORUM DISEASE 15 Another, and perhaps more serious reason, however, was brought out by the report, in 1927, of experiments conducted simultaneously by B. A. Beach, Halpin, and Lampman (21) and by the writers. (22) In these experiments, repeated agglutination tests were made on the same fowls. It was found that many of the fowls that reacted at one test did not uniformly react to subsequent tests. Similar observa- tions have been reported by other investigators including Fitch and Lubbehusen (23) and Newsom, Cross, and Ufford. (24) In the experiments of the writers, a group of birds have been tested at monthly intervals during a period of two and one-half years. Very few of those that reacted positively did so at every test. In some instances, there were as many as from ten to fifteen negative reactions interspersed between two positive reactions. Some of the fowls ceased reacting entirely. Bacteriologic examination of many of these latter gave negative results, indicating probably that they had become free from the infection. Well-established lesions from which Bacterium pullorum has been isolated were found in a number of fowls, however, that died or were killed and that had given a negative reaction to from three to nine tests next preceding their death, thus indicating that their failure to react was not due to absence of infection, but rather to absence of sufficient agglutinins to cause a reaction. It would seem, from the foregoing, that the amount of agglutinins in the blood of a fowl that is a carrier of Bacterium pullorum is sub- ject to rapid and wide variation, and also, therefore, that the per- centage of such fowls that will react to any one test is somewhat uncertain. This is undoubtedly a source of inaccuracy of the test which detracts considerably from its practical value. It is possible that, in the course of time, improved methods of making the test will be found which will increase its efficiency. Proper Application of the Test an Effective Means of Preventing the Disease. — The previously discussed limitations of the test are, to a considerable extent, offset by the high degree of accuracy with which a positive reaction in indicative of infection. Furthermore, there is ample evidence to show that, by the intelligent and systematic appli- cation of the test, Bacterium pullorum infection among both chicks and adults can be sufficiently reduced to warrant its use. Therefore, the agglutination test offers the most effective and practical means for the prevention of pullorum disease available at the present time. Success in such an undertaking requires that all fowls in breeding flocks be tested annually or more frequently; that all reactors be 16 University of California — Experiment Station promptly removed from the premises after each test; that houses, yards, and appliances be kept in a sanitary condition and be given a thorough cleaning and disinfection after each test; that tested stock be not allowed to intermingle with untested stock; and that additions to the flock in the form of hatching eggs, chicks, or grown stock be from known pullorum-disease-free sources. Complete eradication of the disease from a flock will usually require several years when annual testing is the method employed. In some instances this method may serve only to keep the number of infected fowls so small that serious loss occurs infrequently in either the breeding fowls or their progeny. Rapid progress in eradicating pullorum disease from a flock can be made, however, by retesting at intervals of a month or six weeks until none of the fowls react. This can be expected to require from one to five retests. If such a testing procedure is to be adopted, it should preferably be started early enough so that a complete negative test can be obtained before the hatching season starts. Although this practice has not been extensively adopted, the present indications are that it offers a highly efficient means of eradicating pullorum disease from a flock that will probably also prove more economical and satisfactory than annual testing. A flock from Which pullorum disease has in this way been eradicated can thereafter be tested annually. Reasons for Failure to Eradicate the Disease from Flocks That Are Tested. — There are cases of apparent failure of the agglutination test to accomplish much toward the eradication of infection in flocks, which might be cited as evidence that the test was of little value. In many instances, careful investigations would probably reveal that the circumstances were such that failure was to be expected. Hinshaw and Sanders (25) have listed several reasons why the expected results were not obtained from the test. These are based on actual examples that came to their attention during the 1927-28 testing season in Massachusetts and are as follows : 1. Lack of annual testing of 100 per cent of flocks. 2. Failure to promptly remove reactors from the flock. 3. Retaining reactors for egg-producing purposes. 4. Feeding infertile eggs from unknown sources. 5. Failure to properly dispose of offal from reactors killed for home consumption. 6. Buying stock (eggs, chicks, adults) from non-tested flocks. BuL. 486] PULLORUM DISEASE 17 7. Custom hatching for poultrymen who have not tested their flocks. 8. Returning birds to the flock from poultry shows and egg-laying contests without first quarantining and testing. 9. Failure to clean and disinfect the houses following the removal of reactors. 10. Mistakes in numbering blood vials. Application of the Test to Pullets. — In the past it has been con- sidered necessary to delay the testing of pullets until 20 to 30 per cent of the flock were laying, although there seemed to be no exper- imental data to support this theory. The observations of Doyle (27) and Hinshaw and his associates/ 25 ' 26) however, indicate that this is unnecessary. The former obtained a positive reaction to the agglutina- tion test with 11 of 57 two-months old chicks that were survivors of an outbreak of pullorum disease. Hinshaw and his associates tested a small flock (88 birds at the start) five times at intervals of a month. Twenty-one per cent reacted to the first test which was made before any of them had started to lay. At the second test, 3.5 per cent reacted and 10 per cent were laying. At the third test, 4 per cent reacted and 50 per cent were laying. At the last three tests, there were no reactors and all remaining birds were laying. This flock and its progeny, numbering 155 birds, were tested three times about a year later, but no reactors were found. This testing plan was sub- sequently applied to one other flock containing 326 birds. There were 78 reactors to the first test, 5 to the second, 5 to the third, one each to the fourth and fifth, and none to the sixth test. The total number of reactors in the two flocks was 113, of which 97, or 85.86 per cent, were detected at the tests that were made a month before the birds had started to lay. This data, although concerning a small number of birds, suggests that the agglutination test is just as efficient in detecting carriers of Bacterium pullorum when applied to pullets before egg production has started as when applied later. This practice would be desirable from the standpoint of removing infected birds from the flock well in advance of the time that eggs are used for hatching, and would prob- ably be preferred by poultrymen because egg production would not be interrupted to the extent it frequently is when blood samples are taken after production is well under way. 18 University of California — Experiment Station THE RAPID SERUM AGGLUTINATION TEST FOR THE DETECTION OF MATURE FOWLS THAT ARE CARRIERS OF BACTERIUM PULLORUM In 1926, Huddleson and Carlson (28) described a rapid method for performing the agglutination test in the diagnosis of abortion disease of cattle. In this test, small amounts of a concentrated antigen and of blood serum are mixed on a glass slide or plate. The reaction is completed and can be read in from one to fifteen minutes (fig. 4). This rapid method has been applied in testing for pullorum disease Fig. 4. — A positive (left) and negative (right) reaction to the rapid agglutination test. on a fairly large scale in this and other laboratories^ 9 - 30 - 31 « 32 - 33) in comparison with the usual tube or slow method. The results of these comparative tests indicate that the two methods are equally satis- factory in the detection of carriers of Bacterium pullorum; and, it might be added, are probably also subject to the same limitations. The simplicity and speed with which the rapid test may be performed are obvious advantages over the slow test. It must not be thought, however, that the performance of the test is so simple that anyone without previous training or experience can satisfactorily do it. On the contrary, considerable experience is required to correctly interpret the reactions. Veterinarians or others undertaking this work should first spend some time in a laboratory where experience in making and reading tests can be obtained and should have an unquestionably reliable source of good antigen. THE RAPID AGGLUTINATION TEST WITH FRESH WHOLE BLOOD FOR THE DETECTION OF FOWLS THAT ARE CARRIERS OF BACTERIUM PULLORUM In October, 1929, Bunyea, Hall, and Dorset, (34) of the United States Department of Agriculture, Bureau of Animal Industry, described ''a simplified test for pullorum disease" in which whole fresh blood instead of the blood serum was employed. For a test by this method, a BUL. 486] PULLORUM DISEASE 19 drop of blood, obtained by pricking 1 the comb or wing vein, is deposited on a pane of glass and, with a microscope slide, is spread in a thin film over an area of 2 square inches. A drop of antigen is then added to the blood film and mixed with it, either by stirring with a rod or by rocking the pane of glass. The antigen is a heavy suspension of Bacterium pidlorum in physiological salt solution containing no phenol or other preservative. To facilitate reading the results of the test, the glass is placed over a box from which the lid has been removed and which has been blackened inside. Ordinary daylight was found to be adequate unless the work was carried on in a poorly -lighted poultry house; then an electric light was used beneath the glass. A reaction, which, as in all agglutination tests, consists of a clump- ing or flocculation of the bacteria, in the antigen, is said to usually occur in from 5 to 30 seconds but occasionally may be delayed for a minute or more. A fowl is not released until the outcome of the test has determined whether it is a reactor, to be placed in a coop and removed from the flock, or whether it is a non-reactor, to be left in the pen. The test is thus completed and the reactors removed with one handling of the fowls. In conclusion, it is stated that the results of testing a limited number of fowls by this method have been found to check very closely with those of the tube or slow agglutination test. Trials of Agglutination Tests with Fresh Whole Blood at the California Experiment Station. — The simplicity of the procedure of performing the agglutination test with fresh whole blood and the very promising results that had been obtained by Bunyea, Hall and Dorset (34) seemed sufficient to warrant the test being given extensive further trial. Accordingly the writers conducted a series of tests in which the results obtained with the new method and those obtained with the slow or tube method could. be compared. In this series of tests, the rapid serum method was also included. The results of the tests by this method were in very close agreement with the tests by the slow or tube method and, therefore, it is not considered necessary to give them more detailed consideration here. Three thousand seven hundred twenty-seven fowls were used in this work. The results are as follows : Preliminary Trials with 13 Fowls. — These were birds used in other experimental work. Six were known reactors and 7 were known non-reactors to serum agglutination tests. A drop of blood from each was deposited on a glass plate and spread in a film with a micro-slide. 20 University of California — Experiment Station A drop of antigen, prepared according to the recommendations of Bunyea, Hall, and Dorset, (34) was immediately added to the film and stirred with a wooden applicator. Blood samples were taken for the tube agglutination test. For the tube test, dilutions of 1-25, 1-50, 1-100, and 1-200 were used. Seven of the fowls were negative to both tests; 4 gave a positive reaction to the blood film test (see fig. 5) and to the first three or all four dilutions of the slow serum test; 2 Fig. 5. — Upper: Positive reaction to the agglutination test with a drop of fresh blood (left) and a blood film (right). Lower: Negative reaction to agglutination test with a drop of fresh blood (right) and a blood film (left). were negative to the blood film test but gave a positive reaction in the first two dilutions only of the slow serum test. The positive reactions to the blood film test occurred promptly and were distinct. It seemed possible that the two fowls that reacted in the 1-25 and 1-50 dilutions only of the slow serum test failed to react to the blood film test because the amount of blood that became mixed with the antigen was too small. Accordingly, tests of these two birds were repeated using larger and measured amounts of blood. The blood, after being deposited on the glass plate, was not spread into a film but left in the form of a drop to which one drop (.04 cc) of antigen was added. The mixture was then stirred with a wooden applicator and Bui*. 486] PULLORUM DISEASE 21 spread over an area from 2 to 3 cm in diameter. The amount of blood used was .08, .04, .02, .01 and .005 cc. One fowl gave a positive reaction (fig. 5) with the .08, .04, and .02 cc quantities of blood and the other with the .08 and .04 cc quantities. There was no evidence of a reaction with the smaller quantities of blood. The reactions occurred promptly and were just as definite and as easily read as those of tests made with a film of blood or by the rapid serum agglutination method. Additional tests of known reactors and non-reactors to the serum agglutination tests demonstrated that fresh whole blood agglutination tests made with a blood drop could be just as easily and definitely interpreted as those made with a blood film and that tests made with a blood drop would yield more reactors than those made with a blood film. It was, therefore, decided that, in the tests made on poultry farms, both types of fresh blood tests would be used. Trials with Farm Flocks. — Arrangements were made to carry out comparative agglutination tests in two large flocks, one consisting of 2,966 White Leghorns (2843 females and 123 males) and the other of 748 Rhode Island Reds (649 females and 99 males). The procedure was as follows : Two drops of blood from each fowl were deposited separately on a glass plate. One drop of blood was spread into a thin film with a clean micro-slide, the other was undisturbed. One drop of antigen was immediately added to the blood film and one to the blood drop and stirred with a fresh wooden applicator. A reading of reactions was made within 3 minutes. This was considered to be the maximum time it would be practicable to hold the birds if they were not to be released until the results of the tests were known. Further observations were made, however, and reactions that de- veloped slowly but within 7 or 8 minutes were recorded. By that time the blood-antigen mixtures had begun to dry and it quickly became impossible to distinguish with any certainty between changes in ap- pearance of the mixtures caused by drying and those caused by agglutination reaction. A blood sample from each fowl was taken to the laboratory where both slow and rapid serum tests were made. As mentioned previously, the difference between the results of the slow and rapid serum tests was so slight that the details of the rapid serum test results are not included here. The results of the other tests are given in table 2. The data, as summarized in table 2, show that, of the 435 fowls that reacted to the slow serum test, 57.5 per cent reacted to the blood drop test and 39.9 per cent to the blood film test. The number of reactors 22 University of California — Experiment Station to the blood drop and blood film tests would have been decreased by 30 and 24, respectively, had only those reactions been recorded that occurred within 3 minutes. The data further show that the percen- tages of difference between the results of the slow serum tests and the fresh blood tests in the two flocks very closely parallel each other. It is also seen that all fowls that reacted to the blood film test also reacted to both the blood drop and slow serum test and that all fowls that TABLE 2 Comparative Besults of Agglutination Tests of 3714 Fowls by the Blood Drop, Blood Film and Slow Serum Methods Flock number Num- ber of fowls tested Reactors to the slow serum test Reactors to the blood drop test Reactors to blood film test d Reactors to the slow serum test that also reacted to the blood drop test Reactors to the slow serum test that also reacted to the blood film test Num- ber Per cent a Num- ber Per cent a Num- ber Per cent a Num- ber Per cent b Num- ber Per oent b 1 2 Totals 2966 748 3714 262 173 435 8.8 23.1 11.7 150 e 1008 250 5 13.3 6.7 108 f 66 h 174 3 6 8.8 4.6 150 100 250 57.0 57.8 57.5 108 66 174 40 4 38 39.9 a Per cent of number of fowls tested. b Per cent of number of reactors to the slow serum test. c All fowls in this classification also reacted to the slow serum test d All fowls in this classification also reacted to the blood drop and the slow serum tests. e More than 3 minutes were required for the occurrence of 25 of these reactions. f More than 3 minutes were required for the occurrence of 21 of these reactions. g More than 3 minutes were required for the occurrence of 5 of these reactions. h More than 3 minutes were required for the occurrence of 3 of these reactions. reacted to the blood drop test also reacted to the slow serum test. This indicates that the interpretations of the changes in appearance of the fresh-blood-antigen mixtures as positive reactions were both consistent and correct. The analysis of data not included in table 2 shows that : Of 78 fowls that were positive to the slow serum and blood drop test but negative to the blood film test, 48 gave a reaction in either the first three or all four dilutions and 28 gave a reaction in only the first one or two dilutions of the slow serum test. Of 185 fowls that were positive to the slow serum test and negative to both the blood drop and film tests, 69 gave a reaction in the first three or all four dilutions and 116 gave a reaction in only the first one or two dilutions of the slow serum test. From this summary, it is seen that the fowls that reacted to the slow serum test and not to the fresh blood test or that reacted to the blood drop and not to the blood film test were not confined to those that BUL. 486] PULLORUM DISEASE 23 gave a positive reaction in only the 1-25 or 1-50 dilutions of the slow serum test. It cannot be said, therefore, that the failures of these fowls to react to all tests were because the proportion of blood to the antigen was too small. Neither, in the writers' opinion, were they due to errors in reading the test results because the artificial lighting and other facilities for reading that were provided were such that the difference in the appearance of a positive and a negative reaction was very distinct. If it can be assumed that the slow serum agglutination test pro- vides a reasonably accurate means of detecting fowls with pullorum disease, the results of this trial of the fresh blood agglutination test on nearly 4,000 fowls are to be considered as very poor. In its present state of development, therefore, the rapid agglutination test with fresh whole blood does not appear to be a satisfactory substitute for the serum agglutination tests for pullorum disease now in use. Since, as stated previously, the antigen used for the fresh blood agglutination test contains no phenol or other germicidal agent, it is a suspension of living organisms. Its use in making tests on poultry farms is, therefore, attended with some danger of infecting healthy fowls. Another possible criticism of an unpreserved antigen for use outside of the laboratory is that it can easily become contaminated with other organisms. This was found to be the case with a sample of antigen for use in fresh blood agglutination tests that was obtained from a commercial laboratory. THE INTRADERMAL OR PULLORIN TEST FOR THE DETECTION OF MATURE FOWLS THAT ARE CARRIERS OF BACTERIUM PULLORUM An intradermal method of testing fowls for pullorum disease was proposed by Ward and Gallagher (35) in 1917. The test consisted in injecting a small amount of bouillon, in which Bacterium, pullorum has been grown for several weeks, into the skin of the wattle. A reaction is indicated by the presence of a swelling of the wattle in which the pullorin was injected (fig. 6). For several years following, there was very little activity along this line. Interest in the subject was revived, however, in 1926 and 1297 when Graham and his associates (36 > 37) reported promising results from the use of the intradermal test, and when the intradermal diag- nostic agent, termed 'pullorin,' was placed on the market by some manufacturers of biologic preparations. The results of the ' pullorin 24 University of California — Experiment Station test' have since been compared with those of the agglutination test and with post-mortem and bacteriological examinations of tested fowls by number of investigators (Broerman, (38) Edwards and Hull, (30) Gwatkin, (40) Stafseth and Thorp, (41) Bushenell/ 42 - 43) B. A. Beach, (44 > and Michael and J. R. Beach). (45) In the studies made by these investigators, it has been almost uniformly found, first, that a considerable number of reactors to the ■iff ■ .» Jj W "^^'^V:f: Fig. 6. — Fowl showing a positive reaction to the pullorin test. agglutination test did not react to the pullorin test; second, that a considerable number of reactors to the pullorin test did not react to the agglutination test ; and, third, that the post-mortem and bacterio- logical findings were negative in a high percentage of the reactors to the pullorin test only that were killed and examined. In illustration of the preceding statements, a summary of the results obtained by Michael and Beach (45) will be given : BUL. 486] PULLORUM DISEASE 25 There were used in this work four groups of birds, each of which was tested with one of four types of pullorin, i.e., concentrated, precipitated, cell suspension, and cell solution. 5 The concentrated, precipitated, and cell-suspension types of pull- orin are the ones that have been commonly employed in studies of the pullorin test, the first named to a less extent than the other two. The use of the cell-solution type of pullorin has not, to the writers' knowledge, been elsewhere reported. In making the tests, from 0.05 to 0.10 cc of the pullorin was injected intradermally into the wattle. The fowls were examined for reactions twenty-four hours later. After the completion of the test, each fowl was killed and subjected to a bacteriological examination. Those from which Bacterium pullorum was isolated are termed 'bae- teriologically positive;' those from which Bacterium pullorum was not isolated are termed 'baeteriologically negative.' The index of the accuracy of the pullorin reactions are the bacteriological findings, not merely gross lesions or reactions to the agglutination test, as has been frequently the case. The results are given in table 3. The data, as given in table 3, show a wide discrepancy between the results of the pullorin tests, with the exception of those obtained with cell-solution pullorin, and the bacteriological findings. This dis- crepancy consists, first, of the failure of from 7.1 to 55.5 per cent of the Bacterium pullorum infected birds to react to the pullorin test, and, second, of a positive pullorin reaction of from 13.8 to 43.1 per cent of the non-infected birds. Variations in these respects are seen in the results obtained with the concentrated, precipitated, and cell- suspension pullorin, but in all instances the degree of error is sufficient to make the results very unsatisfactory. The results obtained with cell-solution pullorin, however, are in remarkably close agreement s Following is a brief description of the four pullorins. Details of the methods of preparation are given in another publication^ 45 ) Concentrated pullorin. Bouillon cultures of Bact. pullorum incubated for two months, killed by heat, filtered through cotton, and evaporated to one-tenth of the original volume. Diluted with an equal amount of sterile 0.5 per cent phenolized salt solution for use in intradermal tests. Precipitated pullorin. An alcohol precipitate of concentrated pullorin. For the tests a 20 per cent solution of the powdered precipitate in sterile 0.5 per cent phenolized salt solution was used. Cell suspension pullorin. A killed suspension of 48-hour agar cultures of Bact. pullorum in sterile salt solution. Cell solution pullorin. An aqueous solution of Bact. pullorum obtained by alternately freezing and thawing a suspension of the organisms in distilled water about 30 times. The organisms were then removed by passing the suspension through a Seitz filter. The filtrate, a clear slightly amber-colored liquid, was used for the intradermal tests. 26 University of California — Experiment Station with the bacteriological findings. 6 These results, while encouraging, cannot be considered the true value of this product because of the small number of fowls on which it has been used. All of the 566 fowls that were tested with pullorin were also tested by the agglutination method. A positive agglutination reaction was obtained with all but 3 (1.4 per cent) of the 204 birds from which Bacterium pullorum was isolated. This is an unusually high percent- age of reactions to a single test of a group of infected fowls, Of the TABLE 3 Comparison of the Eesults of the Pullorin Test and Bacteriological Examination of the 570 Fowls Num- ber of fowls tested Bacteriologically positive fowls 8 Bacteriologically negative fowls b Type of pullorin Total number That the pullorin test Total number That reacted to the pullorin test Detected Did not detect Num- ber Num- ber Per cent Num- ber Per cent Per cent d 67 353 65 85 9 169 14 15 4 140 13 15 44.4 82.8 92.8 100 5 29 1 55.5 17.1 7.1 0.0 58 184 51 70 8 40 22 1 13 8 21 7 43 1 1 4 a Fowls from which Bacterium pullorum was isolated. b Fowls from which Bacterium pullorum was not isolated. Per cent of bacteriologically positive fowls. Per cent of bacteriologically negative fowls. 362 bacteriologically negative fowls, all except 11 (3.0 per cent) were negative to the agglutination test. This can be regarded as a normal occurrence. The positive agglutination reaction of these bacteriologi- cally negative birds might have been due either to recovery and immunity of the birds or to infection which was present but which was not obtained in culture. The evidence just presented, as well as that obtained by pre- viously mentioned investigators, indicates that the intradermal pullo- rin test in its present status is not a satisfactory means of detecting carriers of Bacterium pullorum. Since, in some cases, a close agree- 6 One hundred forty-three additional fowls have been tested with cell-solution pullorin and by the agglutination method. One reacted to the pullorin test. None reacted to the agglutination test. Bacteriological examination of the one reactor was negative. The other fowls were not available for autopsy. These results provide additional evidence that the cell-solution type of pullorin is not liable to produce a non-specific reaction in healthy fowls. BUL. 486] PULLORUM DISEASE 27 ment has been found between the results of the pullorin test and of the bacteriological examination of the fowls, it is probable that a satisfactory agent for testing by the intradermal method will even- tually be produced. At the present time, however, the systematic application of the agglutination test to breeding fowls, combined with proper management and sanitation in the laying and breeding house, in the brooder house and in the hatchery, provides the most effective and practical means for the prevention and eradication of pullorum disease in adult fowls and chicks. SUMMARY 1. The term 'pullorum disease,' suggested by the Pennsylvania State Department of Agriculture as a substitute for 'bacillary white diarrhea' is accepted as a more appropriate designation for infection with Bacterium pullorum and is used in this publication. 2. Pullorum disease in chicks is an acute septicemic disease, the symptoms and lesions of which are usually not sufficiently character- istic to enable a definite diagnosis to be made without a bacteriological examination. 3. Pullorum disease in adult fowls usually consists of localized infection and lesions in some organ, most frequently the ovaries, and seldom produces visible symptoms. The disease in hens, however, may seriously affect the production, fertility, and hatchability of eggs. 4. Eggs laid by hens with pullorum disease may contain Bacterium pullorum. Chicks hatched from such eggs are infected. This is the most common source of the infection in chicks. 5. Chicks may also acquire pullorum disease from contaminated incubators, brooders, and other equipment. 6. Pullorum disease can be rapidly disseminated among healthy chicks from infected chicks in the same incubators, shipping boxes, or brooders. 7. Pullorum disease in adult fowls can result from their continuing to harbor infection acquired when they were chicks. Infection may also be acquired by mature fowls from consumption of contaminated food, such as infected eggs, or from association with infected adult fowls. 8. Drugs and chemicals have not been demonstrated to have any curative value for chicks actually infected with pullorum disease. Some chemicals, however, particularly the hypochlorite solutions, are of value as disinfectants in the drinking water for chicks to prevent the spread of infection through this route. 28 University of California — Experiment Station 9. The systematic application of the serum-agglutination test to breeding fowls and the prompt removal of reactors, combined with proper management and sanitation on the farms and in the hatcheries, provides the most effective and practical means for the prevention of pullorum disease in adult fowls and chicks. The slow or tube method and the rapid method for performing the agglutination are equally reliable. 10. Repeated agglutination tests at intervals of from 1 to 2 months offer a reasonably certain means of eradicating pullorum disease from a flock in one season. It is believed that owners of infected breeding flocks would find this system of testing more worth while than annual testing. 11. Any testing procedure to be successful must be accompanied by strict adherence to sanitary measures. 12. The agglutination test made with whole fresh blood was not found to be a satisfactory substitute for the slow or rapid agglutination tests made with blood serum. 13. The intradermal or 'pullorin' test in its present status is not a satisfactory means of detecting adult fowls with pullorum disease. LITERATURE CITED 1 Pennsylvania State Department of Agriculture. 1928. Cited by Eettger. 2 Eettger, L. F. 1929. The need of accepted scientific standards and rigid adherence to them in pullorum disease control. Jour. Amer. Vet. Med. Assoc. 74:453-461. s Anonymous, 1929. Conference of laboratory workers in pullorum disease (B. W. D.) eradication. (Mimeo.) News Release, p. 1-3. 4 Bunyea, H., and W. J. Hall. 1928. Some observations on the pathology of bacillary white diarrhea. Paper presented at the annual meeting of the IT. S. Live Stock Sanitary Assoc, s Doyle,, L. P., and F. P. Mathews. 1928. The pathology of bacillary white diarrhea. Indiana Sta. Bui. 323:1-16. fig. 7. g Stafseth, H. J., and E. P. Johnson. 1927. Atypical symptoms and lesions occurring in chicks harboring Bacterium pullorum. Jour. Amer. Vet. Med. Assoc. 71:471-472. 7 Bunyea, H. 1927. An enzootic salpingitis of pullets with special reference to Salmonella pullora infection. Jour. Amer. Vet. Med. Assoc. 70:645-651. BUL. 486] PULLORUM DISEASE 29 s Baudette, F. K. 1929. F. E. Baudette answers Lackie, Poultry Tribune. 3:6, 7, 28, 29. 9 Hinshaw, W. E., H. M. Scott, and L. F. Payne. 1928. Further studies on dissemination of Salmonella pullorum infec- tion in incubators. Jour. Amor. Vet. Med. Assoc. 72:599-610. io Burnett, E. L. 1928. Transmission of Bacterium pullorum infection among mature chickens. Cornell Vet. 18:135-149. ii May, H. G., and H. E. Segelix. 1926. The effect of chemicals in the control of poultry diseases. I Preliminary experiments with bacillary white diarrhea. Poul- try Sci. 6:36-41. 12 Cruickshank, G. A., and II. G. May. 1928. The effect of chemicals in the control of poultry diseases. II An attempt to control fowl typhoid and bacillary white diar- rhea with certain hypochlorite solutions. Poultry Sci. 7:156-162. is Gwatkin, E. 1926. Salmonella pullora infections. Ontario Vet. Col. Eept. 1925-26: 39-48. 14 Coon, C. II. 1928. Some experiments in disinfecting incubators with formaldehyde. Jour. Amer. Vet. Med. Assoc. 72:627-630. i5 Payne, L. F. 1929. B. W. D. control in incubators found at Kansas State College. Hatchery Tribune. 3:1 and 16. is Dakan, E. L., and F. Speer. 1929. Sanitation in the hatchery. Ohio State Agr. Ext. Bui. 90:1-16. 17 Jones, F. S. 1913. The value of the macroscopic agglutination test in detecting fowls that are harboring Bacterium pullorum. Jour. Med. Ees. 27:481-485. is Mathews, F. P. 1926. Obscured reactions in the agglutination test for bacillary white diarrhea. Jour. Immunol. 11:499-504. is Malum an, W. L. 1927. An improved antigen for the agglutination test in bacillary white diarrhea. Jour. Amer. Vet. Med. Assoc. 71:600-606. 20 Eeport of Committee on Poultry Diseases. 1927. Proc. 30th Ann. Meeting U. S. Live Stock San. Assoc. Jour. Amer Vet. Med. Assoc. 70:920-925. 2i Beach, B. A., J. G. Halpin, and C. E. Lampman. 1927. Besults of white diarrhea investigations. Jour. Amer. Vet. Med. Assoc. 70:605-609. 22 Beach, J. E. 1927. Variations in the reactions obtained in repeated agglutination tests of the same fowls with Bacterium pullorum. antigen. Hil- gardia 2:529-544. 23 Fitch, C. P., and E. E. Lubbehusen. 1928. The agglutination test as applied to bacillary white diarrhea. Cornell Vet. 18:19-27. 30 University of California — Experiment Station 24 Newsom, I. E., F. Cross, and O. C. Ufford. 1928. On the accuracy of the agglutination test for Bacterium pullorum infection as shown by repeated tests on the same birds. Jour. Amer. Vet. Med. Assoc. 72:611-617. 25 Hinshaw, W. E., and E. F. Sanders. 1928. Control of Salmonella pullora infection. Massachusetts Agr. Exp. Sta. Control Series. Bui. 43:1-23. 26 Hinshaw, W. R., E. F. Sanders, and G. L. Dunlap. 1929. Eradication of pullorum disease in Massachusetts. Massachu- setts Agr. Exp. Sta. Control Series. Bui. 48:1-36. 27 Doyle, T. M. 1925. Bacillary white diarrhea of chicks. Comp. Path. Therap. 38:266-282. 28 Huddleson, I. F., and E. R. Carlson. 1926. A rapid method for performing the agglutination test in the serum diagnosis of Bang's abortion disease. Jour. Amer. Vet. Med. Assoc. 70:229-233. 29 Beach, J. R. 1927. California Agr. Exp. Sta. Rpt. 1926-27:98. 30 Runnells, R. A., C. J. Coon, and F. Thorp. 1927. An application of the rapid method agglutination test to the diagnosis of bacillary white diarrhea. Jour. Amer. Vet. Med. Assoc. 70:660-662. 3i Stafseth, H. J., and F. Thorp. 1928. Studies of the agglutination and pullorin tests for bacillary white diarrhea. Jour. Amer. Vet. Med. Assoc. 72:745-756. 32 Sawyer, C. E. 1928. Bacillary white diarrhea. West. Washington Agr. Exp. Sta. Rpt. 1927-28:46-47. 33 Bushnell, L. D., and C. A. Brandley. 1929. A comparison of the pullorin reaction and the agglutination test for bacillary white diarrhea. Jour. Amer. Vet. Med. Assoc. 74:364-371. 34 Bunyea, H., W. J. Hall, and M. Dorset. 1929. A simplified agglutination test for pullorum disease. Jour. Amer. Vet. Med. Assoc. 75:408-410. ss Ward, A. R., and B. A. Gallagher. 1917. An intradermal test for Bacterium pullorum infection in fowls. U. S. Dept. Agr. Bui. 517:1-15. (Professional paper.) 36 Graham, R., and E. A. Tunnicliff. 1927. Studies in the diagnosis of bacillary white diarrhea. Jour. Amer. Vet. Med. Assoc. 70:612-627. 37 Graham, R., E. A. Tunnicliff, and E. C. McCulloch. 1927. The intradermal test for bacillary white diarrhea. Illinois Agr. Exp. Sta. Rept. 1926-27:127-132, figs. 1-2. 38 Broerman, A. 1927. Biological and medicinal agents for poultry. Jour. Amer. Vet. Med. Assoc. 70:597-604. ss Edwards, P. R., and F. E. Hull. 1927. A comparison of the agglutination test and the intradermal test in the detection of bacillary white diarrhea. Jour. Amer. Vet. Med. Assoc. 71:590-599. BUL. 486] PULLORUM DISEASE 31 40 GWATKIN, E. 1927. A comparison of the agglutination and pullorin tests for the de- tection of carriers of 8. pullora. Ontario Vet. Col. Ept. 1926-27:42-45. 4i Stafseth, H. J., and F. Thorp, Jr. 1928. Studies of the agglutination and pullorin tests for bacillary white diarrhea as to the efficiency of each in detecting car- riers of Salmonella pullorum infection. Jour. Amer. Vet. Med. Assoc. 72:745-756. 42 BlJSHNELL, L. D. 1928. A comparison of the serologic and pullorin tests for bacillary white diarrhea. Jour. Inf. Dis. 43:60-66. 43 BushnelIi, L. D., and C. A. Brandlet. 1929. A comparison of the pullorin reaction and the agglutination test for bacillary white diarrhea. Jour. Amer. Vet. Med. Assoc. 74:364-371. « Beach, B. A. 1928. Bacillary white diarrhea. Vet. Med. 23:339-342. 45 Michael, S. T., and J. E. Beach. 1929. An experimental study of tests for the detection of carriers of Bacterium pullorum. Hilgardia. 4:185-200. 14m-l,'30