PERFORMANCE OF MORGAN HORSES UNDER SADDLE 5 
tests were made in May. All animals were given approximately the 
same training. The horses were tested at their normal gaits without 
being pushed or allowed to lag. The test at the walk was given before 
the test at the trot with a 2-minute rest period between. The cross- 
country test was given on another day. 
Speed and endurance were determined by trained observers— 
speed, by the time required to cover the total distance in each test; 
endurance, by scores for condition (signs of fatigue) at the end of the 
cross-country ride. The possible range of scores for fatigue was from 
1 to 5, 1 indicating most fatigue and 5 the least. Ease of riding was 
measured by the riders’ scores for ease of handling, performance of 
gait, and ease of gait to rider, at the end of the cross-country ride, for 
the walk, trot, and canter. These scores ranged from 1 to 5, 1 being 
the poorest and 5 the best. 
Other data obtained in the tests and studied for association with 
performance were humidity, temperatures, riders, years, and the 
number of strides on the second, fifth, and eighth laps of the mile 
tests, from which the average length of stride was calculated. 
Studies of association with performance were also made of the fol- 
lowing factors: Height at withers, depth of chest, height at floor of 
chest, distance from point of shoulder to point of hip, heart girth, 
circumference of fore cannon, hind cannon, and knee, width and depth 
of fore cannon and depth of hock, general conformation, style and 
beauty, head, neck, top line of withers, top line of back, top line of 
croup, slope of shoulder, size and shape of feet, quality, condition 
(fleshing), temperament, action at walk, and action at trot. For 
most of these characters a score of 1 was the least desirable and a 
score of 9 the most desirable, but for some characters, such as tempera- 
ment, a medium score was most desirable. 
The data were analyzed principally by the analysis-of-variance 
technique given by Fisher (2) and Snedecor (10). 
RESUETS 
SPEED 
The time required to walk a mile averaged 15.1 minutes (table 2), 
the range among individuals being from 10.0 to 18.5 minutes. The 
variations by years are shown in table 38. Analyses of variance 
showed a significant association between the time required and the 
following: Sires (table 2), years (table 4), length of stride (table 5), 
slope of shoulder (table 6), action at the walk (table 7), riders’ scores 
for performance at the walk (table 8), and size and shape of feet 
(table 18). 
It was impossible to evaluate accurately the effects of sires and 
years on the time required to walk a mile since the same sires were not 
used each year (tables 1 and 3). However, the variance between the 
offspring of sires within the same years and same riders was significant 
table 4). 
Significant differences were found between the time required to 
walk a mile by the offspring of individual dams, but it was not possible 
to separate these effects from those of the sires, owing to the small 
number of offspring from each dam. On the average, there was not 
