The method used by Guo Shou-jing et al to find the times of the solstices from gnomon shadow measurements is discussed. The mean scatter between Guo's measured values of the Sun's shadow and the theoretical values (mean error for a single observation) is found to be between ±1 mm and ±4 mm, and the accuracy of a derived time of the Winter solstice is aobut 0.01 day. The causes for the errors are analysed and a preliminary investigation is made on the derivation of the value of the length of the tropical year adopted in the Shoushi Calendar.