Reactivity of peripheral blood leukocytes against human foetal cells. II. Cytotoxic potential of preparations enriched or depleted of different leukocyte populations

Leukocytes separated initially by Ficoll-Triosil sedimentation from the peripheral blood of 62 healthy donors and 64 patients with a variety of cancers were tested for cytotoxicity against cells from a single human foetal lung, by means of a microplate technique based on visual enumeration of surviving target cells. Under these conditions cytotoxicity was primarily observed to be a function of the granulocyte content of the preparations, which was extremely variable (range 0-60%). When the influence of these cells was greatly diminished through depletion by adherence, residual cytotoxicity was still detectable in the majority of samples tested. Complete abrogation of this cytotoxicity by pretreatment with carrageenan implicated cells of the monocyte-macrophage series, which persisted as a small minority (about 2%) in the already depleted effector cell populations. However some cytotoxicity could also be demonstrated in populations, purified by passage through columns containing nylon fibre or Degalan-coated (human Ig-anti Ig) beads, which consisted almost exclusively (about 98%) of lymphocytes, mainly of thymus (T)-dependent type (greater than or equal to 80%). Under these conditions and those in which T-cells were concentrated by formation of spontaneous E rosettes, cytotoxicity appeared to decline with T cell enrichment, suggesting that the cytolytic effects were attributable to cells of the non T-compartment which were present to a variable extent in the different T-cell-enriched populations. Although the absolute identity of these cells was not established, the experiments illustrate potential sources of non-specific cytotoxicity among different effector cell populations against target cells expressing antigens to which the majority of leukocyte donors--in health or disease--were presumably non-sensitized. As such these data have several implications for the interpretation of in vitro cytotoxicity tests against human neoplasms.

Erythropoietic inhibitory activity of plasma fractions from hypertransfused sheep

The plasma and serum of polycythemic animals and man are reported to exhibit erythropoietic inhibitory activity in certain bioassay systems. The plasmas of hypertransfused and normal sheep were fractionated by the methods of Cohn and Weimer and their associates, and the major fractions assayed for stimulatory or inhibitory activity in the exhypoxic polycythemic mouse assay system. The results indicate that the inhibitory activity acquired with hypertransfusion can be demonstrated in fraction VI of Cohn's and in precipitate of D of Weimer's methods and that the activity of each is retained in the eluates from G-200 Sephadex columns.

IgM pyroglobulinemia with erythrocytosis presenting as hyperviscosity syndrome. I. Clinical features and viscometric studies

The hyperviscosity syndrome is described in a patient with erythrocytosis and an immunoglobulin M with kappa light chain (IgMK) macroglobulinemia. Viscometric studies were carried out on whole blood and demonstrated the contribution of both the increased hematocrit value and the macroglobulinemia to the whole blood viscosity. Clinical improvement followed phlebotomy and was accompanied by a decrease in whole blood viscosity. Continued treatment with chlorambucil has been associated with a long symptom-free period. The macroglobulin was characterized as a monoclonal IgMK pyroglobulin which retained its thermoprecipitability was reduced to 7S monomers. The presence of IgMK aggregates in the serum may have contributed to the increase in blood viscosity.

Aplastic anaemia: Evidence for an immunological mechanism

The soft agar culture assay (C.F.U.-C) has been used in vitro as a measure of haemopoietic capacity of bone-marrow. In a patient with aplastic anaemia pretreatment of the patient's bone-marrow with horse anti-human-thymocyte globulin and complement (A.T.G. + C) prior to culture led to a dramatic increase in ability to form colonies in the soft agar assay; and co-culturing marrow from a normal donor and from the patient resulted in a distinct reduction in the number of expected C.F.U.-C. These findings point ot an immunological or autoimmune mechanism in this patient by selective destruction of the suppressing cells in the patient's marrow with A.T.G. and by suppression of normal myelopoiesis following addition of the patient's marrow to normal marrow.

Inhibition of transplanted sarcomas mediated by BCG in rats with a defined immunological deficit

Experiments were undertaken to test the hypothesis that a major component of BCG contact-induced inhibition of syngeneic tumour growth in rats is not dependent on the participation of thymus-processed (T) cells. Hosts were deprived of T cells by thymectomy followed by either lethal irradiation (850 rad) and bone marrow reconstitution, or repeated whole body irradiation to a total dose of 1,000 rad. After 6 weeks had elapsed to allow for bone marrow restitution, rats were challenged with trypsinized sarcoma cells admixed with Glaxo strain BCG. For sarcoma P7, host T-cell deprivation did not significantly diminish the capacity of BCG to prevent the progressive development of this neoplasm from an inoculum of one million cells. Under similar conditions, the incidence of sarcoma CC5 development in maximally deprived hosts was significantly greater (7/19) than in normal controls (1/16) (P is less than 0.05), but the majority of rats (58%) did not succomb to tumour outgrowth. In the case of a third neoplasm--a spontaneously metastasizing fibrosarcoma (P8)--the effect of BCG on primary tumour development was comparable in normal and deprived recipients and was limited to temporary arrest as distinct from complete inhibition. Assessment of the influence of BCG on lung metastases was more complex since the extent of metastatic disease from subcutaneous tumour cells alone was greater in deprived rats than in normal rats. It is concluded that T-cell participation is not a major requirement for BCG contact-induced inhibition in this system and some implications for the mechanism of action are discussed.

Leucocytotoxicity in malignant and non-malignant colonic diseases

The survival of cells from five different cultures of allogeneic malignant colonic carcinoma, two from normal adult colonic epithelium and eight from foetal colonic epithelium in the presence of leucocytes from patients with neoplastic and inflammatory disorders of the colon has been compared. Cytotoxicity assessed by the reduction of the number of adherent target cells in microplate wells compared with those surviving in wells treated with tissue culture medium alone was observed with leucocytes from donors in all categories examined including those from individuals without any known abnormality. Patients with ulcerative colitis were the only group to reveal consistent reactivity against cultures derived from all three sources, an observation which may reflect sensitization to organ-related antigens in this disease. In contrast, leucocytes from patients with bowel neoplasia showed reactivity for cells derived from colon carcinoma tissue, which was comparable to that of healthy donors. Evidence for tumour-specific cytotoxicity was therefore lacking in this study. It is suggested that the detection of tumour-associated antigens on cultured cells may be limited by a number of factors of which the wide variation in reactivity among controls and unspecified nature of the target cells are likely to be of greatest importance.

Fatal disseminated adenovirus infection in a renal transplant recipient

A 61 year old woman died of diffuse interstitial adenovirus pneumonia 55 days after receiving a cadaveric renal allograft. The adenovirus was serologically distinct from the 33 known human adenovirus serotypes and appears to represent a new human adenovirus. Pathologic and virological findings indicate that the pneumonia was only one manifestation of a disseminated infection, the source of which may have been a latent adenovirus infection preexisting in the donor kidney. The establishment of the etiologic diagnosis in this case, which was complicated by the presence of oculocutaneous and esophageal herpes simplex virus infection as well as focal pulmonary aspergillosis, required coordinated histopathologic and virological investigation. Our findings demonstrate that severe viral infections in transplant recipients are not caused exclusively by members of the herpesvirus group.

PHA stimulation of separated human lymphocyte populations

Lymphocyte preparations from peripheral blood and tonsils were separated into populations enriched with T or B cells by formation of rosettes with SRBC and separation of the rosette-forming and non-rosette-forming populations. T cell-enriched populations were also prepared by nylon column filtration. Using these methods preparations were obtained which comprised 80--95% T or B lymphocytes as determined by E-rosette formation and surface immunoglobulin (Ig) staining. PHA responsiveness, measured by [3H]thymidine incorporation, varied between relatively wide limits and was critically dependent on the degree of separation obtained. Relatively pure B-cell populations (less than 12% T cells) from blood and tonsils gave low PHA responses while preparations from blood still containing 24--38% T cells gave responses equal to or even greater than those of unseparated controls (60--78% T cells). T cell-enriched populations (80--86% T cells) responded to an equal or greater degree than controls but more efficient separation (greater than 90% T cells) resulted in markedly reduced stimulation. There was thus no simple correlation between the degree of phytomitogen-induced transformation and the number of T cells present. It is concluded that the low response of relatively pure T-cell populations may be due to depletion of B cells or non-lymphoid cells (or both) during the separation procedures. These observations have implications for the use of PHA stimulation as a measure of T-cell activity in mixed populations such as those of human peripheral blood leucocytes.

Tumour inhibition mediated by BCG in immunosuppressed rats

Two rat sarcomas (CC5 and P7) which grow progressively on transplantation into normal syngeneic hosts failed to develop when injected in admixture with the Glaxo strain of Bacillus Calmette-Guérin (BCG). Under comparable conditions, the local development of a third neoplasm (P8) was temporarily inhibited and the number of pulmonary metastases significantly reduced. Experiments were undertaken to determine the extent to which the anti-tumor action of BCG required an immunocompetent host. Rats were immunosuppressed by sub-lethal whole-body irradiation (450 R), with or without prior thymectomy and challenged with inocula of mixed BCG and tumour cells when their capacity to respond to bacterial, tumour and unrelated antigens was maximally depressed. In non-sensitized immunosuppressed rats, the ability of BCG to limit tumour outgrowth was abrogated only in the case of sarcoma CC5. For this neoplasm, immunogenic in syngeneic hosts by conventional criteria, there was a statistically significant difference in the number of tumours in immunosuppressed rats (51%) compared with non-sensitized immunocompetent controls (6%). Presensitization to either bacterial or tumour antigens, prior to thymectomy and/or irradiation, fully restored the tumour-inhibitory capacity of BCG. By contrast, sarcoma P7 was not significantly less susceptible to BCG-induced regression in non-sensitized immunosuppressed rats than in nonsensitized normal rats; and sarcoma P8 similarly failed to reveal any significant differences in susceptibility to BCG affecting primary or secondary tumour development. It is concluded that tumours may vary widely in their sensitivity to host reactions aroused by BCG. Certain neoplasms, exemplified by sarcoma CC5, require participation of an immune reaction of delayed hypersensitivity type for optimal destruction at BCG sites, while for others (e.g. sarcoma P7) an immunoreactive component of this type is not essential. By contrast, a third category of tumour (e.g. sarcoma P8) is relatively resistant to host reactions induced by the mycobacteria. An important component of BCG-mediated tumour inhibition is not dependent on an immunologically intact host.

Cell surface antigen expression on chemically induced murine leukaemias

The immunogenicity of murine leukaemias induced by chemical carcinogens or irradiation in C57Bl or (C57Bl times DBA2) F1 hybrid mice has been studied in vivo by transplantation and in vitro by indirect membrane immunofluorescence (IF) using syngeneic immune or allogeneic immune antisera. Two of 5 leukaemias tested for immunogenicity by assessment of the capacity of syngeneic mice specifically immunized with irradiated (3 Krad) cells to reject small challenge inocula (10(3)-10(4) cells) displayed weak neoantigenicity while 3 were non-immunogenic by this criterion. Antibodies directed against cell-surface antigens of the immunizing cells of 7 leukaemias were not detectable by immunofluorescence tests using sera from the respective immunized mice. H-2 histocompatibility antigens readily identified on normal lymphoid cells using reference Balb/c anti-C57Bl (H-2d anti-H-2b) alloantisera could neither be detected on the majority of transplanted leukaemias nor on 9 primary leukaemias in C57Bl mice induced by N-butyl-N-nitrosourea (BNU). Two of the transplanted leukaemias showed greatly diminished capacity for absorption of alloantibody compared with normal spleen cells. Transplantation to H-2 different recipients, in which the leukaemic cells were invariably rejected, generated a strong humoral antibody response, which was demonstrable against normal lymphoid cells. Failure to demonstrate significant antibody binding by indirect immunofluroescence tests with immune sera, or by absorption, is presented as evidency that H-2 antigen expression is substantially modified on BNU induced leukaemia cells. These findings have implications for the detection of tumour neoantigens on chemically induced leukaemias.

Cell-mediated cytotoxicity to human pulmonary neoplasms

Peripheral blood leukocytes from patients with confirmed pulmonary neoplasia were tested for cytotoxicity against cultured cells derived from lung tumours of various histological types, foetal and normal adult lung tissue and tumours arising in organs other than the lung. Leukocytes from 73 percent of patients were cytotoxic for lung-tumour derived cells compared with age- and sex-matched normal donors, while the frequencies of reactivity against normal adult lung-derived cells and cells from unrelated tumours (e.g. bladder, colon, breast) were 42 percent and 18 percent respectively. Leukocytes from lung cancer patients were also cytotoxic for cells derived from foetal lung but susceptibility to cytolysis was variable, cells from 13- and 14-week embryos revealing greatest reactivity (88 percent). Leukocytes from patients with a variety of tumours of non-pulmonary origin or with non-malignant conditions (including respiratory disorders) were also reactive with lung-tumour-derived target cells but with a lower overall frequency (35 percent) than those from lung-cancer patients. The significance of these cytotoxicity data for the existence of tumour-specific host immunoreactivity in lung neoplasia is discussed.

Inhibition of mitogen mediated lymphocyte blastogenesis by adenosine

The effect of adenosine on the proliferative response of human peripheral circulating lymphocytes to stimulation by concanavalin A, phytohemagglutinin and pokeweed mitogen was evaluated. Increasing concentrations of adenosine substantially inhibited mitogen mediated lymphocyte blastogenesis. Erythro-9(2-hydroxyl-3-nonyl) adenine. HCl enhanced the inhibitory effect of adenosine. Inosine, the deamination product of adenosine, had an inhibitory effect which was less than that of adenosine. Inhibition by adenosine may be relevant to the normal regulation of immune function and may account in part for the pathophysiological relationship between severe combined immunodeficiency disease and adenosine deaminase deficiency.

Studies on the mechanisms of chemical leukaemogenesis

Following a single injection of MNU into "intact" mice, a high incidence of leukaemia (90%) is obtained, with a 50% induction time of 200 days. Immunological studies indicate that the θ antigen is expressed on the leukaemic cells. Thymectomized MNU treated mice had a 50% induction time of 500 days, and the incidence was somewhat lower. Leukaemias failed to develop in MNU treated T lymphocyte deficient animals and in lethally irradiated, or thymectomized lethally irradiated mice reconstituted with MNU treated bone marrow. It is suggested that the T lymphocytes rather than the haemopoietic stem cells or pre-T cells are the "target cells" in MNU leukaemogenesis.

Antigens of tumours induced by naturally occurring murine sarcoma virus (MSV-FBJ). II. Detection of cell-surface antigens by indirect membrane immunofluorescence

Cell surface antigens expressed by cells transformed in vivo by FBJ virus, a wild type murine sarcoma virus (MSV) complex derived from a spontaneously arising sarcoma in a CF1 mouse, have been studied by indirect membrane immunofluorescence (MIF). Using mouse antisera raised by immunization of syngeneic CBA mice with transplanted FBJ sarcomata an antigen common to all FBJ tumours was detected which was also present on Gross (G) antigen positive tissues, viz. leukaemic and preleukaemic AKR lymphoid cells, but absent from the tissues of mice of G negative strains. Failure to demonstrate antigenic cross-reactivity in reciprocal MIF tests using FBJ immune sera and antisera to MSV-H (Harvey), an MSV isolate of Friend-Moloney-Rauscher (FMR) sub-group specificity, established the virus type-specificity of antigens expressed by sarcoma cells transformed by the respective MSV.The presence of a cellular antigen with G specificity on FBJ sarcoma cells was confirmed in tests with aged exbreeding C57B1 antisera containing naturally occurring G antibody lacking significant virus neutralizing activity. However, evidence for a "sarcoma-non-leukaemia" antigen on cells transformed by MSV-FBJ was not obtained since absorption studies failed to reveal any specificity on FBJ sarcoma cells which was not also present on AKR leukaemic tissues.It is suggested that the major humoral component of the immune response to FBJ sarcoma cells is evoked against antigens specified by the associated non-pathogenic leukaemia virus (MLV-FBJ) and the relationship of antigens demonstrated by MIF to those detected previously by complement fixation (CF) and tumour rejection tests is discussed.

Antigens of tumours induced by naturally occurring murine sarcoma virus (MSV-FBJ). I. Detection of group and type specific antigens by complement fixation

Antigens associated with cells transformed in vivo by FBJ virus, a wild type murine sarcoma virus (MSV) complex originating from a spontaneously arising osteosarcoma in a CF1 mouse, have been partially characterized by complement fixation (CF). Using rat antisera against antigens specified by Gross leukaemia virus (GLV) the group specific (gs) antigen of C-type RNA murine tumour viruses was demonstrated in FBJ tumours as well as in GLV rat leukaemias, AKR lymphomata and sarcomata induced by MSV-H (Harvey), an MSV isolate of Friend-Moloney-Rauscher (FMR) subgroup specificity. Using mouse antisera against antigens present in FBJ cells the Gross (G) or wild type specificity of FBJ tumours was demonstrated by cross reactivity with antigens expressed on normal AKR lymphoid tissues and leukaemias. These antigens were absent from MSV-H induced sarcomata and in reciprocal tests mouse antisera to MSV-H failed to react with antigens present in FBJ tumour cells. No distinction between cellular and virion antigens expressed by FBJ cells was possible by CF although evidence for a cellular antigen with G specificity was obtained in tests using aged C57B1 antiserum containing a naturally occurring G antibody lacking significant virus neutralizing capacity. However, the likelihood that mouse FBJ antisera contain antibodies to type specific viral envelope antigens (VEA) as well as cellular antigen is discussed.

Immunogenicity of rat hepatoma membrane fractions

The principal expression of immunity elicited in syngeneic rats immunized with rat hepatoma membrane fractions was the development of a tumour specific antibody response. This antibody was demonstrable by membrane immunofluorescence staining of viable hepatoma cells in suspension and the sera exhibited complement dependent cytotoxicity for cultured hepatoma cells. In the absence of complement, however, membrane immune sera were highly "blocking", protecting plated hepatoma cells from attack by sensitized lymph node cells. The cell mediated immune response elicited by hepatoma membrane immunization was weak, as evaluated by the colony inhibitory activity of lymph node cells for hepatoma cells in vitro or the adoptive transfer of immunity with peritoneal exudate cells. Correlated with this overall pattern of immune response, membrane immunization did not elicit tumour rejection reactions. These findings are relevant to current views that humoral factors operate antagonistically to limit cell mediated immunity to tumours. A further relevant feature was the observation that membrane immunization, eliciting a prominent humoral immune reaction, conditioned the recipients so that they subsequently failed to elicit a tumour rejection immunity on treatment with irradiated tumour cells.

Circulating antibodies in human connective tissue malignancy

In a comparative study, sera from patients with connective tissue tumours, various carcinomata and from individuals without malignancy were evaluated by indirect immunofluorescence (IF) for antibodies reactive with apparently specific antigens shared by sarcoma derived tissue culture cell lines; for antibodies by IF to 2 tissue autoantigens (nuclear antigen and smooth muscle antigen) on rat liver substrate; and for HL-A antibodies by microcytotoxicity against a panel of 22 lymphocytes.

Antibodies reactive with 11/16 cell lines originating from sarcomata were detected in 36% of all sarcoma sera tested, 12% carcinoma sera and 9% sera from controls. The incidence of antisarcoma antibody (ASA) was higher in the sera of sarcoma patients whose disease was in the primary phase (53%) compared with those in whom disease was advanced. A lower incidence (8%) of antibodies to nuclear antigen was detected in the sera of sarcoma patients compared with carcinoma patients (22%) and controls (23%), but the incidence of smooth muscle antibodies (SMA) was higher (45%) compared with carcinomata and controls (36% and 35% respectively).

HL-A antibodies were present in 13% sarcoma sera, 36% carcinoma sera and 33% control sera.

Evidence is presented to show that the various antibodies are distinct and that those reacting with sarcoma derived cell lines may be tumour associated antibodies possibly related to a virus specified antigen.

Tumour-associated transplantation antigens of neoplasms induced by a naturally occurring murine sarcoma virus (FBJ-MSV)

FBJ osteosarcoma virus (FBJ-MSV) isolated originally from a spontaneously arising osteosarcoma in a CF1 mouse is the only known naturally occurring murine sarcoma virus (MSV). It is unique among strains of MSV in producing primarily sarcomata in mice. The capacity of tumour cells transformed in vivo by this agent to elicit specific transplantation immunity in syngeneic hosts was investigated. A low level of resistance (10(4)-10(5) cells) was consistently induced by implantation of x-irradiated (15,000 rad) tumours or surgical excision of developing subcutaneous grafts. By contrast intraperitoneal inoculation of virus containing cellfree extracts of FBJ-MSV sarcomata was a far less effective immunization procedure. Confirmatory evidence for the antigenicity of these neoplasms was obtained in tests in which preincubation of tumour cells with lymphoid cells from specifically immune donors inhibited in vivo outgrowth of the FBJ-MSV cells in untreated syngeneic recipients. The induction of host resistance to FBJ-MSV cells by immunization with identical and independently-induced FBJ-MSV tumours established that FBJ-MSV cells possess common cell surface antigenic specificities in a manner analogous to those of experimental neoplasms induced by other oncogenic DNA and RNA viruses. Since FBJ-MSV cells release infectious virus it was not possible in this system to establish whether the tumour-rejection antigen was cellular or virion in nature. The antigenic weakness of FBJ-MSV cells in syngeneic hosts is comparable with that of virus-induced murine leukaemias of the Gross (G) or "wild" type subgroup to which category FBJ-MSV also belongs. These features suggest that FBJ-MSV exemplifies naturally occurring sarcomagenic viruses more closely than those of the Friend-Moloney-Rauscher-Graffi (FMRGr) subgroup which in general induce highly antigenic neoplasms.

Studies on the antigenicity of radiation-induced murine osteosarcomata

The immunogenicities of 15 murine osteosarcomata induced with a bone seeking radioisotope ((90)Sr) in normal and chimaeric CBA mice were studied. Attempts were made to induce tumour-specific immunity in syngeneic mice by treatment with x-irradiated (15,000 rad) tumour or surgical excision of developing subcutaneous tumour grafts. Resistance was evoked against 6 tumours and this was relatively weak. With the remaining tumours, no resistance against the immunizing tumour could be demonstrated, even though the transplantation tests were made highly sensitive by the use of inocula of as few as 2 × 10(3) cells in pre-irradiated (400 rad) hosts. Sera from mice immunized against each of the tumours were tested against viable cells of the immunizing tumour by indirect immunofluorescence. In no instance did tumour antisera give a convincing reaction with tumour cells although alloantisera raised by hyperimmunization of H-2 identical and H-2 different donors with osteosarcomata consistently gave strongly positive reactions. The results are interpreted as illustrating the weak tumour specific antigenicity of radiation-induced murine osteosarcomata. The possibility that antigenic deficiency is a consequence of immunosurveillance in this tumour system is discussed.

FBJ virus-induced tumours in mice. A histopathological study of FBJ virus tumours and their relevance to murine and human osteosarcoma arising in bone

Nine of the 15 neonatal CBA mice injected intramuscularly with a Moloney concentrate containing FBJ virus developed tumours: likewise 5 of 7 CBA neonates injected intraperitoneally with a cell-free filtrate derived from a transplanted tumour of the former group.Of soft tissue origin, these FBJ sarcomata have a characteristic histological appearance and are of low grade malignancy. Although occasional islets of cartilage osteoid and bone were noted, these were regarded as indicative of evolutionary metaplasia in the collagenous matrix of pleomorphic fibroblastic sarcoma. No tumour was acceptable as osteosarcoma of conventional type and osseous origin. There were, however, additionally 2 minute spindle cell sarcomata arising in femoral periosteum and non-neoplastic periosteal proliferation was observed. The differences of these FBJ fibroblastic sarcomata from murine osteosarcoma-either spontaneous or induced by Sr 90-are emphasized. Furthermore, their deviation from the structural pattern and behaviour of human osteosarcoma is discussed.

Respiratory Exchange and Body Size in the Aldabra Giant Tortoise

1. The O2 consumption and CO2 release of nine giant tortoises Testudo gigantea (weight range 118 g-35·5 kg) were measured at a temperature of about 25·5°C. Four European tortoises Testudo hermanni (weight range 640 g-2·16 kg) were also used. The mean RQ values obtained were 1·01 for T. gigantea and 0·97 for T. hermanni. These values were not influenced by activity or size.

2. The data was analysed by plotting log/log regression lines relating body weight to O2 consumption. Both maximum and minimum metabolic rates recorded for each individual T. gigantea showed a negative correlation with body weight. For active rates the relation was O2 consumption = 140·8W0·97, whereas for inactive animals O2 consumption = 45·47W0·82.

3. The maximum rates were obtained from animals that were observed to be active in the respirometer and the minimum rates from animals that remained quiet throughout. The scope for activity increased with body size, being 82 ml/kg/h for animals of 100 g and 103 ml/kg/h for 100 kg animals. The corresponding ratio between maximum and minimum rates increases from about 2 to 6 for the same weight range.

4. Values for metabolic rate in T. hermanni seem to be rather lower than in T. gigantea. Analysis of the relative proportion of the shell and other organs indicates that the shell forms about 31% of the body weight in adult T. hermanni but only about 18% in T. gigantea of similar size. The shell is not appreciably heavier in adult T. gigantea (about 20%).

5. Data obtained for inactive animals is in good agreement with results of other workers using lizards and snakes. Previous evidence suggesting that chelonians show no reduction in metabolic rate with increasing size is not considered to conflict with data obtained in the present work.