A method of circulating and sampling the rumen contents of sheep fed on ground, pelleted foods.

Considerable difficulty and uncertainty can be associated with sampling through cannulas from the reticulo-rumen of the sheep. The contents are by no means homogeneous, and for some purposes relatively large samples have to be withdrawn from various places within the organ and mixed in an external container before a subsample is taken for analysis and the unused portion returned to the rumen. In addition to the uncertainties of this conventional method of sampling, it is a rather cumbrous procedure and rumen contents may be lost from the open cannula if the animal bleats or moves rapidly. Mixing of materials added to the rumen is a slow process. J. G. Gordon (1960, personal communication) has studied the concentration of polyethylene glycol (PEG) in samples withdrawn as described above and has found it to fluctuate widely for several hours after introduction of the marker. The rumen contents are in fact a relatively unhomogeneous and slowly mixing pool. The early stages of experiments that depend on tracing the concentration or specific activity of materials added to the rumen have therefore an inherent uncertainty. Examples of experiments particularly sensitive to this uncertainty are those involving measurement by isotope dilution of short-chain fatty acid production (Sheppard, Forbes 8t Johnson, 1959 ; Gray, Jones & Pilgrim, 1960). Finely ground, pelleted diets are becoming of increasing importance in the nutrition of ruminants and, being stable and easily stored and handled, they are useful for certain types of experimental work. Although theoretically with such diets a small sample can be made representative of the whole, their effect in decreasing rumen motility can increase the difficulties of experiments when sampling from the rumen is involved. The rumen contents remain heterogeneous on these diets, the proportion of dry matter being higher at the surface than in the lower part of the rumen. We thought, however, that with these diets continuous artificial mixing of the rumen contents could be achieved and that added materials would be rapidly mixed with the rumen contents. This paper describes the construction and use of a pump which withdraws rumen contents from a more or less central position in the rumen, forces them through an external circuit containing a sampling device before returning them to the liquid surface in the rumen, and thus gives rapid mixing of rumen contents. Experiments