Search

Background. Rodent models that accurately reflect human filovirus infection are needed as early screens for medical countermeasures. Prior work in rodents with the Zaire species of Ebola virus (ZEBOV) primarily used inbred mice and guinea pigs to model disease. However, these inbred species do not show some of the important features of primate ZEBOV infection, most notably, coagulation abnormalities. Methods. Thirty-six outbred guinea pigs were infected with guinea pig–adapted ZEBOV and examined sequentially over an 8-day period to investigate the pathologic events that lead to death. Results. Features of disease in ZEBOV-infected outbred guinea pigs were largely consistent with disease in humans and nonhuman primates and included early infection of macrophages and dendritiform cells, apoptosis of bystander lymphocytes, and increases in levels of proinflammatory cytokines. Most importantly, dysregulation of circulating levels of fibrinogen, protein C activity, and antifibrinolytic proteins and deposition of fibrin in tissues demonstrated both biochemical and microscopic evidence of disseminated intravascular coagulation. Conclusions. These findings suggest that the outbred guinea pig model recapitulates ZEBOV infection of primates better than inbred rodent models, is useful for dissecting key events in the pathogenesis of ZEBOV, and is useful for evaluating candidate interventions prior to assessment in primates.

The polymerase chain reaction method (PCR) was used to investigate events in the pathogenesis of varicella-zoster virus (VZV) infection in strain 2, Hartley, and euthymic hairless guinea pigs. VZV was detected in peripheral blood mononuclear cells (PBMC) obtained 2-5 days after infection in 8 (50%) of 16 strain 2, 4 (40%) of 10 hairless, and 10 (34%) of 29 Hartley guinea pigs. The frequency of VZV-infected PBMC was estimated to be at least 1/200,000, which is comparable to that observed in human infection. When VZV PCR was used to test ganglia from hairless guinea pigs, samples from 6 of 8 animals were positive. Of 45 VZV-infected guinea pigs that were tested for cellular immunity by VZV T lymphocyte proliferation assay, 44 developed a stimulation index > 2.0. Control animals had no detectable virus by PCR and did not develop cellular immunity to VZV. These experiments showed that viremia was detectable by PCR during primary VZV infection of guinea pigs in about half of the animals regardless of the strain of guinea pig. Acquisition of cellular immunity provided a consistent marker of infection in all guinea pig strains. PCR was also useful for demonstrating VZV in guinea pig ganglia tissue, with VZV gene sequences being detectable for at least 80 days after infection. With the combination of PCR and immunologic assays, various guinea pig strains should be useful for studies of VZV pathogenesis and for the evaluation of antiviral agents and vaccine strategies.

An animal model permitting study of congenital infections with cytomegalovirus (CMV) has been developed in guinea pigs. Fifteen Hartley strain guinea pigs in the latter half of pregnancy were inoculated intraperitoneally with 10(5.5) 50% tissue culture infective doses of guinea pig CMV. Forty percent of infected mothers delivered litters containing at least one infected newborn, as defined by a positive culture of lung, spleen, or brain. All tissues were cultured by an explant technique. The three mothers who had no detectable complement-fixing antibody to CMV prior to experimental infection delivered infected litters, whereas three of 12 immune mothers delivered infected litters (P less than 0.01). A low-passage, tissue culture-adapted virus produced neonatal infection as frequently as did salivary gland-passaged virus. No congenital abnormalities were found in any of the seven infected newborns. CMV was isolated from lung, spleen, or brain in the four newborns of nonimmune mothers; CMV was isolated from lung only in the three newborns of immune mothers. These preliminary experiments demonstrate that the guinea pig is a suitable animal for further study of maternal-fetal CMV infections.

The effects of pregnancy on primary acute and chronic persistent infection with cytomegalovirus (CMV) were examined in the guinea pig model. Pregnant guinea pigs developed a more severe acute CMV infection than did nonpregnant animals. High death rate, pronounced splenomegaly, interstitial pneumonia, and necrosis with inclusions in various tissues were observed only in pregnant animals. Acutely infected pregnant animals also differed from nonpregnant animals by the absence of tissue inflammation and by a delay in the peripheral lymphoproliferative response. During persistent infection, salivary gland virus titers were significantly higher in pregnant animals than in nonpregnant guinea pigs. These results indicate that during acute guinea pig CMV infection, virus clearance and recovery from clinical disease are altered in the pregnant host. Furthermore, enhancement of persistent chronic guinea pig CMV infection occurs in the salivary glands during pregnancy.

We developed an enzyme-linked immunosorbent assay (ELISA) for measuring the antibody response to herpes simplex virus (HSV) glycoprotein D (gD). The ELISA was specific and more sensitive than immunoblotting techniques. Antibody to HSV gD was detected in 10 of 12 sera (geometric mean titer, 36.3) obtained from HSV-2-infected guinea pigs 14 days after intravaginal inoculation, and the titer increased to greater than or equal to 5,120 on day 60. Therapy with acyclovir delayed and diminished the antibody response to gD, although by day 60 the titers of antibody to gD in acyclovir-treated animals were not significantly different from those in controls. An increased titer of antibody to gD seemed to be associated with a reduced number of recurrent episodes and to a reduced number of days with recurrent lesions.

Artificial mixtures of the glycoproteins B, C, D, and E of herpes simplex virus 1 and 2 (HSV-1 and HSV-2), purified individually from infected Vero cell lysates by immunoaffinity to monoclonal antibodies, were bound to an aluminum hydroxide gel and were used to immunize mice, guinea pigs, and owl monkeys (Aotus trivirgatus) once or twice (mice and guinea pigs) to as many as four times (owl monkeys). In all animals tested, low levels of neutralizing antibodies were detected only after two or more immunizations. Lymphocyte transformation tests in owl monkeys suggested low or borderline levels of cellular immunity. The survival of immunized mice after intracerebral challenge was inversely related to the challenge dose. Immunized guinea pigs challenged by intravaginal inoculation showed reduced morbidity at the site of inoculation and were protected from CNS disease. Both immunized and nonimmunized monkeys were highly susceptible and could not be differentiated with respect to morbidity or mortality when challenged with 1,000 pfu of HSV-2 by the intravaginal route.

Guinea pigs inoculated with the live M-44 vaccine strain of Coxiella burnetii were examined grossly and microscopically for the presence of Q fever-related lesions. Mild myocarditis was observed in 38% of the infected animals but in none of the control animals. Livers showed significant incidence of hepatitis, necrosis, and granuloma formation, especially during the first eight days of the infection. A much lower incidence of splenitis was also found but was considered to be of borderline significance. Generally, lesions were mild in nature, and none seemed to endanger the life of the animal or to cause observable distress.

Human recombinant interleukin-2 (rIL-2) modifies infection with herpes simplex virus type 2 (HSV-2) in normal guinea pigs. Animals were injected sc with rIL-2 twice a day, beginning 24 hr before infection and continuing for three subsequent days. Guinea pigs were assigned to five different regimens in which rIL-2 was administered daily at dosages from 8 X 10(3) U/kg to 8 X 10(5) U/kg. After intravaginal inoculation with HSV-2, 83% of 24 control animals developed apparent genital herpes, and 5% had asymptomatic viral shedding. Animals receiving 4 X 10(4) U or 2 X 10(5) U or rIL-2/kg showed a significantly lower rate of infection (P less than .001 by chi 2) and a lower number of lesions and more rapid healing than did the infected animals in the control (untreated) group. The mortality in the untreated group (28%) was higher than in the animals in the 2 X 10(5) U/kg treatment group (7%). No difference was observed between the control, the 8 X 10(3) U/kg, and the 8 X 10(5) U/kg treatment groups with respect to severity of infection or mortality. Animals examined two to four weeks after inoculation had lymphocyte stimulation indices (induced with HSV-2 antigen) greater than 3.5 and titers of antibody between 2.5 log10 and 4.5 log10 by radioimmunoassay. In the disease-free animals, stimulation indexes were less than 2 and antibodies were undetectable. The mean stimulation index was 54 +/- 16 in the diseased control animals and 12 +/- 7 (P less than .05) in the 2 X 10(5) U/kg treatment group.(ABSTRACT TRUNCATED AT 250 WORDS)

Treatment of cutaneous Herpesvirus hominis type 2 infection in albino guinea pigs with 8-methoxypsoralen plus long-wave ultraviolet light was evaluated by clinical assessment and quantitative viral cultures. There was a statistically significant difference between the treated animals and the controls in relation to vesicular development, rate of healing, and viral titers. The treatment was effective after the appearance of lesions, when viral multiplication has already begun.

Most commonly used antiamoebic drugs are effective in invasive amebiasis, but their response against trophozoites of Entamoeba histolytica, present in the lumen of the human colon, is inadequate. We report the development of an antiamoebic drug carrier that may be effective against luminal infections. Our preparation consists of small silica particles (5-10 microns in diameter) covalently linked to a potent antiamoebic drug, 2-(4-aminophenoxymethyl)-5-nitro-1-methyl imidazole. Silica-drug particles were injected into mice, hamsters, and guinea pigs. We found that trophozoites phagocytosed the particles in vivo and in vitro, followed by rapid cell death due to the released drug. Analysis of mouse serum revealed that no drug was absorbed from the intestine after placement of the drug-containing particles in the intestine. The antiamoebic activity of particles recovered from the intestine was almost fully retained. This novel antiamoebic concept may be useful for luminal therapy for asymptomatic amebiasis and may minimize side effects and frequency of administration.

The attenuated M-44 strain of Coxiella burneti was investigated for biological characteristics affecting its use as a live vaccine. After 25 serial passages in guinea pigs, the strain did not develop phase I properties detectable by the microagglutination test. Infectious organisms were not detected in untreated guinea pigs later than 12 weeks after inoculation; however, infection could be reactivated at later times by the production of stress (by pregnancy or drug treatment) in guinea pigs that had recovered. Evidence of reactivated infection was observed in recovered pregnant animals just before parturition at 46 weeks after infection, in methyl prednisolone-treated animals at 56 weeks, and in cyclophosphamide-treated guinea pigs at 96 weeks. A low rate of intercage and intracage infection was recorded.

Thirty-four clinical isolates of Sporothrix schenckii were studied for their temperature sensitivity and morphology in vitro. Four selected isolates were also studied for their morphology and growth rate in vivo. Conidia from 26 clinical isolates obtained from lymphocutaneous or extracutaneous lesions formed colonies at both 35 C and 37 C in vitro within five days. Conidia from eight isolates obtained from fixed cutaneous lesions (nonlymphangitic) formed colonies at 35 C but failed to do so at 37 C during a two-week incubation period. Studies of the growth rate at 25 C and morphology in vitro did not show any clear differences between the two groups of isolates. When the isolates that grew at 37 C were injected into mice by the intraperitoneal route, the organisms multiplied well in the internal organs but multiplied even better in the testes. Mice developed skin lesions on paws and tails within 72 days after injection. When the isolates that did not grow at 37 C were injected into mice by the same route, the organisms did not multiply in the internal organs but grew well in the testes. These mice did not develop any skin lesions during the 72day observation period. Epidermal or intradermal inoculation of conidia into guinea pigs showed that the isolates capable of growth at 37 C produced skin lesions with a lower concentration of conidia than the isolates not growing at 37 C. Of four additional isolates from natural sources, three grew at 37 C in vitro. Cutaneous sporotrichosis is a chronic mycosis caused by Sporothrix schenckii. The most common method of acquisition is traumatic implan

A guinea pig model of immunosuppression was utilized to study the effects of immunosuppressive chemotherapy on lung response to challenge with Pseudomonas aeruginosa. Study groups included normal guinea pigs, as well as guinea pigs that received a one-week course of cortisone acetate (CA, 100 mg/kg per day) plus 15 mg of cyclophosphamide (CTX)/kg per day (CA + LoCTX group) or 30 mg of cyclophosphamide/kg per day (CA + HiCTX group). Separate groups received CA or HiCTX alone. Intratracheal instillation of P. aeruginosa resulted in bilateral hemorrhagic pneumonia in both normal and immunosuppressed animals. Survival was 100% for normal animals and for those given CA alone, 67% in the CA + LoCTX and the HiCTX groups, and 0 in the CA + HiCTX group. Increased mortality correlated with a diminished polymorphonuclear leukocyte inflammatory response in infected lung tissues and also with the addition of CA to CTX. Clearance of viable P. aeruginosa from lung tissue was significantly reduced in animals receiving the combination CA + HiCTX. Thus, decreased lung inflammation and the addition of CA appeared to be important determinants for fatal pseudomonas pneumonia.

Single, large, daily aminoglycoside doses in animals are less toxic than conventional dosing, and higher drug concentrations in vitro produce more-rapid bacterial killing. Thus, we compared various aminoglycoside dosing schedules in neutropenic (n = 153) and nonneutropenic (n = 192) guinea pigs with Pseudomonas aeruginosa pneumonia. Equivalent tobramycin dosages were given: 5 mg/kg every 4 h or 30 mg/kg every 24 h. Animals were serially killed during therapy, and quantitative lung cultures were performed. Bacterial titers in lungs dropped rapidly in all tobramycin-treated animals, both neutropenic and nonneutropenic, during the initial 16 h of therapy. In nonneutropenic guinea pigs, lung titers remained constant despite continued 4-h dosing. With subsequent 24-h dosing, titers continued to drop, and by 72 h there were a significant number of animals with sterile lungs (P less than .01). In neutropenic guinea pigs given tobramycin every 24 h, bacterial regrowth occurred; thus, therapy was ineffective. Adding mezlocillin, however, suppressed regrowth; thus, combination therapy was superior (P less than .05).

Guinea pigs vaccinated with bacille Calmette-Guérin (BCG) and unvaccinated guinea pigs were challenged by the respiratory route six weeks or six months after vaccination and sacrificed at various intervals after challenge. The six lobes of the lung were cultured separately, and the percentage of culture-positive lobes was calculated, as well as the log10 number of virulent bacilli recovered. The latter was subjected to an analysis of variance, which compared the fate of bacilli in the four largest lobes with the fate of those in the two smallest lobes. The results indicated no difference between the six-week and six-month intervals between vaccination and challenge. In the longer intervals between challenge and sacrifice, small numbers of secondary lesions could be seen on the lobes of the BCG-vaccinated animals. It was concluded that vaccination with BCG retarded the onset and/or reduced the extent of hematogenous dissemination of virulent mycobacteria to the lungs.

The epidemic of equine encephalomyelitis, reported by Meyer, Haring and Howitt1 in 1931 as caused by a filtrable virus, made its appearance during the summer and autumn of 1930 in certain rural districts of California and continued to recur during the corresponding periods of 1931 and 1932. The disease also spread into Nevada 2 and the northwestern states and reached Nebraska in the fall of 1932.

have not been determined although both in man and in animals a complete loss of capacity for skin reaction has been frequently associated with overwhelming or terminal tuberculous infection. That this loss of skin reactivity is the rule in the later period of experimental tuberculosis in rabbits has been recently shown by Freund, Laidlaw, and Mansfield.1 Freund2 has also shown that skin sensitivity to tuberculin is slight in young infected guinea pigs compared to that of adult animals, although systemic hypersensitivity as measured by the lethal intraperitoneal dose develops to the same degree in young as in adult animals. Factors affecting the degree of sensitivity other than age, certain intercurrent infections, terminal tuberculous infections, or actual specific desensitization by repeated inoculation have not been demonstrated experimentally. The experiments to be described were undertaken to determine the effect, if any, of induced hypothermia on the course of experimental tuberculosis in guinea pigs. The most conspicuous result was neither a retardation nor an acceleration in the progress of the infection, as might have been anticipated, but a very definite retardation in the development of skin sensitiveness.

Dehydrated cultures of tubercle bacilli suspended in olive oil promptly loose their acid-fastness upon incubation, and the subcutaneous or intraperitoneal injection of these non-acid fast cultures into guinea-pigs sensitizes to the intracutaneous administration of old tuberculin.1 The purpose of this paper is to present a series of results obtained by treating guinea-pigs with the non-acidfast cultures before and after inoculating them with living tubercle bacilli.

Attempts to define mechanisms of host susceptibility to Shigella infection would be facilitated by the development of a suitable small animal model. The natural resistance of common laboratory animals to experimental infection with shigellae has been overcome by a number of methods. Successful infection of mice and guinea pigs can be accomplished by pretreatment with antibiotics to alter intestinal flora sufficiently to minimize its protective effects [1]. However, such infections are almost invariably asymptomatic and thus are of limited value for the study of natural disease. Fatal Shigella infection can be produced in adult guinea pigs by starvation or subcutaneous injection with carbon tetrachloride, provided that calcium carbonate is given before and opium administered after oral challenge [2, 3]. Germfree guinea pigs are also susceptible to shigellosis [4]. Injection of substances adversely affecting the Krebs cycle prior to oral challenge has been reported to induce fatal infection in mice [5]. Studies in guinea pigs indicate that an essential step in the pathogenesis of shigellosis is penetration of the intestinal mucosa by infecting organisms [6]. Since cellular turnover in the intestinal mucosa operates at an exceedingly rapid rate

Occasionally during the past, some of the guinea-pigs used in this laboratory were observed to have large, fluctuant swellings of various groups of lymph nodes. Animals suffering from this disease, known as epizootic lymphadenitis or "lumps," had been found in colonies of stock guinea-pigs and in groups of animals used in various experiments. In some instances, the character of the thick or caseous pus led to a superficial diagnosis of tuberculosis or pseudotuberculosis. No detailed attention, however, was paid to the condition until several guinea-pigs, primarily affected with "lumps" died of an acute infection with an hemolytic streptococcus shortly after they had been inoculated with tubercle bacilli, the spirochete of rat-bite fever or other organisms. During a period of six months, following September, 1928, a search for guinea-pigs with this disease was made in a colony averaging about 150 animals. In this period, 20 guinea-pigs with "lumps" were found. From all of these the same unusual type of mucoid hemolytic streptococcus was isolated. The organism was obtained in pure culture from 16 of the animals and in 4 cases only was it associated with staphylococci. As the animals from which mixed cultures were obtained had draining sinuses from ruptured abscesses, it is probable that the staphyloco.cci were secondary invaders, entering through this channel. The prompt removal of infected guinea-pigs from a group reduced the incidence of the disease to a minimum.

The abilities of attenuated smooth (strain 19) and rough (strain 45/20) Brucella abortus to immunize guinea pigs were tested 20 weeks after vaccination by respiratory challenge with virulent (strain 2308) B. abortus. The two vaccine strains were presented as living organisms in a small dose (106 organisms) or as killed organisms in a large dose (1010 organisms) in Freund's incomplete adjuvant. All vaccines gave significant protection; the superiority of living strain 19 was demonstrable only at a low level of challenge. In parallel studies with different animals, vaccinated guinea pigs were tested for delayed hypersensitivity. Animals given all the vaccines had comparable skin reactions after intradermal injection of brucellergen. All of the vaccines primed animals for the production of IgG against smooth antigen following skin-test with brucellergen. Long-lasting immunity to infection was conferred only by vaccines that induced delayed hypersensitivity and primed for antibody production, but it was not possible to dissociate these effects.

A model for the study of delayed hypersensitivity to hepatitis-associated antigen (HAA) in guinea pigs was designed. Guinea pigs immunized with purified HAA plus Freund's complete adjuvant developed a positive skin reaction when they were inoculated intradermally with HAA 17 to 35 days later. This reaction is characterized by erythema and induration that reaches a maximum at 24 to 36 hr and then subsides over the next 48 hr. Histologic examination of the skin lesions shows an infiltrate of predominantly mononuclear cells. Heating HAA for 1 min at 37 C, 57 C, 75 C, or 98 C does not modify its activity as an immunizing or skin-reactive agent.

Since freshly isolated strains of H. pertussis are used to vaccinate the human being against whooping cough, it was thought interesting to determine what protection such strains would bestow upon guinea pigs, animals highly susceptible to this type of organism. The first experiment was planned to determine the difference in protection acquired by guinea pigs that had been actively immunized against a freshly isolated strain grown on blood agar and those that had been actively immunized against the same strain acclimated to and grown on veal brain agar when both vaccinated groups were subsequently injected intraperitoneally with an MLD of a freshly isolated culture. The MLD of the freshly isolated organism was 3 mgm. by wet weight in 1 cc. of saline. The MLD of the same strain acclimated to and grown on veal brain agar was 35 mgm. by wet weight in 1 cc. of saline. The "disease producing" and "fixed" phases' were grown on their respective mediums and both harvested after 2 days and standardized so that 1 cc.-5 mgm. of wet weighed organisms. Fortyeight animals weighing from 210 to 250 grams were used. Thirty-two animals were immunized; 16 with the "disease producing" or freshly isolated and 16 with the "fixed" or old phase of the same organism. Five tenths of a cc. of the standardized vaccine was injected the first day, 1.0 cc. 4 days later and again 1.0 cc. 4 days after that. Miller2 reported protective antibodies at their height within 12 days after the injection of freshly isolated vaccine so that an interval of 12 days was allowed to elapse and then all of the animals including 16 controls were injected intraperitoneally with 1 MLD (3 mgm.) of the freshly isolated or the "disease producing" phase of the organism. Only one animal lived and it had been protected by the fixed strain. Fortyseven of the 48 injected animals died, 39 of them within 3 days after the intraperitoneal injection. The typical reaction seen at the autopsy of guinea pigs that died after injections of freshly isolated organisms was a massive dilatation of the capillaries with splotchy hemorrhages in the parietal abdominal wall, a salmon red color of the small intestines and a free serosanguinous exudate. Of the various groups of animals tested, there was but slight difference in the pathological reactions of those that died. Five of the first and 4 of the second group had atypical reactions as compared to only 2 of the control group. This nu

In a preliminary note on the " Production and Concentration of a Serum for Rocky Mountain Spotted Fever"1 we reported the preparation of a serum for Rocky Mountain spotted fever and its potency, determined by tests on guinea-pigs. Additional experiments have confirmed our previous results and have shown that the serum has a limited curative effect on guinea-pigs inoculated with spotted fever virus, if given during early stages of the disease. A series of experiments upon the efficacy of sodium cacodylate as a curative agent has also been completed, the results of which are reported in this paper.

Through the courtesy of Dr. F. A. Rich, head of the veterinary department of the Vermont Agricultural Experiment Station, I received a number of guinea-pigs from a breed naturally deficient in complement. According to Moore,1 this characteristic property is inheritable. Both Moore and Downing (cited by Moore 1) found that in some instances as much as 1 c c of the serum produced no hemolysis of sensitized corpuscles. Moore also found that the apparent lack of complement in the blood serum of some guinea-pigs is not due to the presence therein of anything interfering with the action of the amboceptor. Coca 2 recently reported that both the mid-piece and the end-piece of the complements of these guinea-pigs are present. "There is lacking only the so-called 'third piece/ addition of inactive guinea-pig or human serum containing this third component to the deficient serum causes hemolysis in about three times the minimal hemolytic quantity of normal guineapig's serum." The third component according to this author is not identical with the lipoid cytozyme (thrombokinase), since the blood of the complement-deficient guinea-pigs clots normally. Having found 2 guinea-pigs, the serums of which failed to hemolyze 0.5 c c of a 5 % suspension of sensitized sheep corpuscles in doses of 0.8 and 1.0 c c, it was thought of interest to find out whether an increase of amboceptor alone will produce activation of the deficient serum. The amboceptor was employed in doses of 2, 10, 50, 100 and 500 units. The unit of cells was 0.5 c c of a 5% suspension of sheep corpuscles. The complement-deficient serums were employed in doses of 0.1 cc. The incubation period was 1 hour at 37 C. Table 1 shows clearly the effect of the increased amboceptor content. Complete hemolysis occurred in the tube containing 500 units of amboceptor. The complement was employed in doses of 0.1 c c, thus 8 and 10 times

Aqueous and phenolic extracts and insoluble residues from smooth biotype 1 strains of Brucella abortus and Brucella melitensis were studied for immunologic and biologic comparison. Brucella differ from enterobacteriaceae in a high glucose content and lack of 3-deoxy-D-mannooctulosonic acid (KDO) in aqueous phases, in a rhamnolipid in insoluble residues, and in the high content of nitrogen in extracts made with acetic acid. Some chemical similarities may exist between Pseudomonas and Brucella. Extracts of Brucella, especially phenol extracts of B. melitensis, induce in adult mice a curable wasting disease. None of these fractions sensitize guinea pigs (as determined by dermal hypersensitivity test), but most brucella fractions are allergens for animals already infected with B. abortus. Most of the differences in biologic behavior and in antibody induction between B. abortus and B. melitensis could be associated with an unidentical distribution of similar compounds among the various fractions. In addition, B. abortus differs from B. melitensis in the lack of both galactosamine and methionine.

The statement has frequently been made that, while the bacillus of bovine contagious abortion (Bacillus abortus, Bang) will cause abortion in other domestic or laboratory animals when experimentally inoculated, it is not known to be epizootic among any animals except cattle. Abortion of an epizootic nature occurs not infrequently with horses, sheep, and swine. In no instance, however, has such a case been connected with the organism found in cattle. There is a considerable body of evidence to show that the natural infection of animals other than cattle with this organism must be exceedingly rare. However, the ease with which other species of animals may be artificially infected leaves it by no means certain that under the proper conditions natural infection might not occur. In this connection, the following observations are of interest. In 1908 the Department of Animal Husbandry of this station began some work with infectious abortion. From that time until the end of 1911 a number of guinea-pigs were inoculated from time to time with material from aborting cows. During the year 1911 a number of these animals were inoculated with pure cultures of the bacillus abortus. During this period the space for keeping inoculated animals was very limited. A room in the station piggery had been fitted up as a growing room for such small animals. During a large portion of the year there was no place suitable for inoculated pigs except in this room. The inoculated pigs were kept in wire cages of the Vaughan type. These cages were on legs about six inches high, and were placed in the growing pens so that the healthy pigs could run under the cages. No inoculated pigs were ever returned to the general herd and none of the healthy animals could come in contact with the infected pigs except through the one-eighth inch wire screen on the sides of the cages. However, such an arrangement allowed infected litter and urine to reach the