2, CIRCULAR 824, U. S. DEPT. OF AGRICULTURE 
lin (4, 4), Jackson (3), and Steele (11). Few attempts have previously 
been made to devise and use objective tests under carefully controlled 
conditions similar to those in actual practice for ordinary saddle 
horses. 
MATERIALS AND METHODS 
Data were available on 79 three-year-old Morgan horses that were 
raised and trained at the United States Morgan Horse Farm, Middle- 
bury, Vt. Eleven of these animals had a tendency to pace and were 
omitted from the study. The remaining 68 horses were sired by 8 
different stallions and ranged from 2 to 25 offspring per sire. There 
Ficure 1.—Morgan mare Damsel being tested under saddle at the trot onthe 
training track at Middlebury, Vt. Time for 1 mile, 4.2 minutes; stride, 12.3 
feet. 
were 6 stallions, 20 geldings, and 42 mares. Most of the horses were 
tested in the spring of the year that they were 3 years old. All data 
were not available on all horses. Distribution of the offspring by 
sires, years, and sexes is shown in table 1. The training and testing 
procedures have been described by Phillips, Speelman, and Williams 
(9) and Phillips (6), together with the general plan of the breeding 
program; hence only the tests under saddle are described here. 
The tests under saddle included a 1-mile walk around an eighth-of- 
a-mile oval track, a 1-mile trot around the same track (fig. 1), and an 
11.35-mile cross-country ride. The track had a surface of sand and 
clay and was smoothed and rolled at frequent intervals during the 
tests. The cross-country ride was on unpaved roads and over rolling 
country with several steep grades. During this ride each horse walked 
4.65 miles, trotted 5.65 miles, and cantered 1.05 miles. Markers 
were placed along the course to indicate to the riders the gaits to be 
maintained over each portion, as shown in figure 2. Each horse 
