Polarized light biomicroscopy was used to examine the normal pre‐corneal tear film in 25 dogs. Blink rates were also determined. The normal range of surface morphology of the canine tear film was established. Three categories of surface patterning of the lipid layer of the tear film were identified. These were: (a) a wave‐like pattern, (b) a pattern of irregular islets, and (c) a pattern resembling the polished surface of granite. All three patterns may occur in the same eye. The inter‐blink thickness of the lipid layer of the pre‐corneal tear film varied between 0·013 μm (the minimum thickness discernable by this method) and over 0·581 μm (the maximum thickness at which coloured interference fringes are clearly defined) in localized areas. The mean total blink rate was 14·2/min. Overall, incomplete blinks constituted approximately 66 per cent of the total blink rate, but the ratio of complete to incomplete blinks varied considerably between animals. By dividing the dogs into thick and thin lipid layer groupings, it was possible to demonstrate a statistically greater rate of complete blinks (P < 0·05, Mann‐Whitney U test) in animals with a thinner lipid layer. The applicability of polarized light biomicroscopy to veterinary clinical ophthalmology is discussed.

Polarised light biomicroscopy was used to examine the normal pre-corneal tear film in 21 eyes of 12 pekingese dogs. The purpose of the study was to examine the influence of excessive exophthalmos on the pre-corneal tear film in the dog. The majority of the animals were found to have high levels of ocular surface contamination by particulate material and plaques of viscous mucus. Other abnormalities included surface lipid with an abnormal granular (three dogs) or 'curdled' (two dogs) appearance; excessive thinning of the lipid layer of the tear film; and the presence of dark globular structures in two dogs, which were presumed to be abnormal meibomian lipid. Break up of the tear film was observed in one dog. Grossly, a thread of viscous mucus was frequently observed along the margin of the lower eyelid. It is postulated that this thread forms because of the excessively exophthalmic conformation of the breed, which prevents the normal access of effete mucus and entrapped debris to the lower conjunctival fornix. The combination of the above factors in the pekingese is suggested as the mechanism whereby the tear film has a reduced stability, thus enhancing the risk from factors more usually considered to initiate corneal ulceration in the breed. The possible adverse effects of lid splitting for the mass removal of distichiae in exophthalmic dogs is discussed.

Polarized light biomicroscopy was used to examine the pre‐corneal tear film in 14 normal cats. Blink rates were also determined. The normal range of surface morphology of the feline tear film was established. Two principal categories of surface patterning of the lipid layer of the tear film were identified. These occurred in all eyes examined and were: (a) a wave‐like pattern, and (b) a pattern of variable character resembling the polished surface of granite. Over most of the area of the tear film, the thickness of its lipid layer between blinks was under 0·089 μm in all cats examined. However, narrow streaks of thicker lipid (up to 0·511 μm) were occasionally observed. The mean total blink rate was 2·4/min. Overall, incomplete blinks constituted approximately 50 per cent of the total blink rate. The tear film of the cat is compared with that of man and the dog.

Polarized light biomicroscopy was used to examine the pre‐corneal tear film in 16 dogs which were suspected of, or confirmed to be, suffering from keratoconjunctivitis sicca. One animal displayed a normal tear film, while the other 15 showed variable degrees of abnormality. The abnormal features included: contamination of the ocular surface with particulate debris and plaques of mucus, discontinuity and excessive granularity of the surface lipid film, pronounced thinning of the aqueous phase, and breakup of the tear film. After parotid duct transposition, the tear film of some animals was almost indistinguishable from normal. However, in others the surface lipid layer was thinned, discontinuous, or virtual absent. Discontinuity of surface lipid was most obvious in animals with marked salivary epiphora, and was associated with a variable amount of crystalline concretion over the ocular surface. The clinical usefulness of polarized light biomicroscopy for the diagnosis, and subsequent monitoring, of keratoconjunctivitis sicca is discussed.

Two sample groups of domestic cats were assessed for (a) central corneal thickness (17 eyes), using optical pachymetry, and (b) horizontal and vertical corneal diameter, using a photographic method (20 eyes). The average corneal thickness was 0.755 +/- 0.33 mm. The mean horizontal diameter was 16.5 +/- 0.60 mm and the mean vertical diameter was 16.2 +/- 0.61 mm. The horizontal corneal diameter was significantly greater than the vertical corneal diameter (p less than 0.001, paired t test).