233mastitis the four quarters were cited as A, B, C and D as follows: A, left front; B, right front; C, left hind; D, right hind. The simplest way to ensure that the claw was correctly applied was found to be the use of a standard position for the bucket relative to the direction in which the cow was facing. The bucket was always placed with the arrow on the lid pointed to the head of the bail, and care was taken that there was no twist in the long connecting tubes to the claw. The claw was applied to the teats with the straight tubes along instead of across the bail. With this system there was no need to alter the position of the bucket for right-and left-hand bails. It was found that there was little difficulty in using the claw as described instead of in the more usual position, transverse to the bail. Milking proceeded normally as for an ordinary bucket plant. The strip bucket was used in the same way with the arrow towards the head of the bail, so giving automatically an adjustment for right-and left-hand bails. On completion of stripping, the two buckets were placed side by side with the arrows pointing in the same direction, and the strippings were added to the corresponding quadrant container in the milking-machine bucket before weighing and sampling.The coliform organisms have probably received more attention than most other groups of non-pathogenic bacteria owing to their importance as ' indicator' species in water analysis, wherein two distinct procedures have long been employed, viz. (a) a presumptive test to demonstrate the presence or absence of coliform species in general, performed by inoculation into bile salt-lactose broth (Britain) or lactose broth (America), and (b) confirmatory tests designed to isolate and differentiate faecal from non-faecal types.Subsequently the 'coliform test' has been widely adopted in Britain for the examination of raw-milk supplies, and quantitative standards have been introduced for grading special designated raw milks. Similar tests are used extensively in pasteurized milk control, although no legal standard has been adopted. In order to facilitate examination of large numbers of samples, only the presumptive test is performed in dairy work. The validity of this procedure has been questioned by numerous workers, particularly in America, on the ground that interpretation of the test must be uncertain since faecal types are not differentiated. Conversely, it is argued that the presumptive test is sufficient since undesirable changes in milk or milk products may be produced equally by faecal and