Inter-rater reliability in reading bacterial counts was investigated by
Drs. Viel and J. Allen of the Ontario (Canada) Veterinary College.
15 animals with respiratory infections each had one swab of their nasal
cavity taken, which were then each applied to one slide. The two raters
each read the 15 slides twice for bacterial count.  No information is given
on whether the order in which the slides were read was randomized.

The data are shown with columns for rater (1 or 2), reading (1 or 2),
slide (1 through 15), and bacterial count.

These data originally appeared in Shoukri and Pause (1999), Statistical 
Methods for Health Sciences, Second Edition, New York: CRC Press, p.34.