Introduction
Whatever method of training is used, the basic principle of progressive overload should be considered (Pickering 1972). That is, ensuring that the work load is set at a slightly higher level than that to which the individual is normally accustomed, and gradually increasing the work load over time. Consideration should also be given to the concept of the specificity of training (Watson 1983 and Fox 1984). Best results usually occur when the majority of training for a particular sport bears a close resemblance to that sport, both in terms of type of training and any equipment used.

For cycle racing the majority of training is done riding a bike on the road or a track. However, weather conditions or other factors may occasionally preclude such training and, if the rider wishes to continue on-the-bike training during this period, an alternative method must be found. Recently, increasing use has been made of inexpensive indoor training devices to provide the necessary work load. This article reports on the results of a series of training sessions for which the subjects attached their own bikes to an ergometer designed by the author (Brooke and Firth 1974), as shown in Figure 1.

Table 1 shows the comparison between this heart rate measurement procedure and heart rate calculated from an electrocardiograph, using Students t-statistic. The statistically insignificant difference between the two methods indicates the validity of manual palpation for these levels of work heart rates.

For the majority of the training sessions monitoring of the heart rate throughout was considered impractical, so they were checked at the following intervals:

• 5min continuous ride - after 3min and again in the last 10sec;
• 5 x 2min rides - in the last 10sec of the third work period and again in the last 10sec of the fifth work period;
• 10 x 1min rides - in the last 10sec of the third, fifth and tenth work periods;
• 10 x 30sec rides - in the last 10sec of the third, fifth and tenth work periods.

Statistical analyses of the improvements in overall mean power output during the training sessions, and in the mean race speeds over 10 miles and 25 miles were carried out. There was a significant improvement of the overall mean power output in the training sessions after three weeks (p ‹ 0.010), six weeks (p ‹ 0.001), nine weeks (p ‹ 0.010) and twelve weeks (p‹ 0.002). The improvement in mean race speed at 10 miles was not significant after three weeks, but was significant after six weeks (p ‹ 0.010), nine weeks (p ‹ 0.010) and twelve weeks (p ‹ 0.010). Improvement in race speed at 25 miles was also not significant after three weeks but was significant after six weeks (p ‹ 0.10), nine weeks (p ‹ 0.050) and twelve weeks (p ‹ 0.010).

Figure 3 gives a graphical illustration of the relationship between the overall mean power output in the training sessions and race performances at 10 miles and 25 miles. The figure shows a strong positive correlation between overall mean power output in the training sessions and performance at 10 miles (r = 0.935, p ‹ 0.001) and 25 miles (r = 0.979, p ‹ 0.001).

Discussion
It is not the purpose of this article to suggest that this training method is superior to other on-the-bike methods. The statement is that this form of training is a valid method which follows the principles of progressive overload and the specificity of training. This can be shown from a number of standpoints.

General. It can be seen from Figure 3 that there were significant improvements in the subjects' ability to tolerate the power output demanded in the training. Similar increases have been elicited in initially untrained subjects exercising on bicycle ergometers (Gleser and Vogel 1971). The present subjects however, were well accustomed to regular training, both immediately prior to this 12-week program and in previous years. During the course of the training period the six subjects recorded between them 38 personal best performances in 10 miles and 25 miles time trials. None had shown such improvement in previous years.

Progressive Overload. As the ability to tolerate a particular work load improves, the heart rate at that load becomes less (Rowell 1964 and Ekblom 1968). Conversely, to keep the work heart rate at a constant level the work load has to be gradually increased over time. This latter statement follows the principle of progressive overload and was adopted as the basis for setting up the training program reported here.

Fitting the subjects' own bikes to a calibrated training device meant that the work load could be set for each individual. It could also be compared with what had been used in previous sessions and with race performances. By keeping the work heart rates within the ranges specified for each section of the training session, it allowed any improvements in the subjects' fitness to be immediately noticed and catered for by a regular increase in the work load. Thus the training system provided a constantly updated record of how each subject was reacting to the training program, vital information for maintaining motivation.

Specificity of Training. Comments are frequently made on the specificity of fitness, the way in which it is attained, and the way it is assessed (Rasch and Morehouse 1957, Hamley and Thomason 1967, and Astrand and Rodahl 1970). The strong relationship in the present study between the training and competitive work levels is not surprising when one remembers that, in both situations, the subjects used their own bikes, to which they were well accustomed.

Intermittent Work. Whilst it is true that a time triallist needs a good cardiorespiratory capacity (Brooke, Hamley and Stone 1970), there is also a need for a high power output. After all, the prizes in such competitions are awarded to the fastest (most powerful) rider. To keep a maximal stimulus on the oxygen uptake and the cardiac output during continuous exercise it is only necessary to work at about 80% of maximal work output (Astrand and Rodahl 1970). Higher intensity work only exhausts the subject sooner without further increasing either the cardiac output or oxygen uptake. However, it seems that a higher work intensity is required to tax skeletal muscle capacity (Nett 1970). By using an intermittent work training method it is possible to use a much higher work load whilst still keeping a strong stimulus on the cardiac output and oxygen uptake (Astrand and Rodahl 1970).

Racing cyclists need to produce a mean power output of approximately 235 watts to maintain a speed of 23mph, and approximately 300 watts to maintain 25mph (Whitt 1971). In the present study the subjects were close to these race speeds at the beginning and end respectively of the 12 weeks training period. Their overall mean power output during the training sessions were 338 watts at the beginning of training (representing 26.25mph) and 418 watts at the end of the training period (representing 28mph). This shows that the use of an intermittent work schedule allowed the subjects to tolerate greater stress in training (albeit for short work periods) than they encountered in their best competitive efforts.

Pedalling Rates. It has been shown that racing cyclists prefer pedalling rates in the range 90-120 revs/min, and even as high as 150 revs/min (Ulmer 1973). Ulmer concluded that power output rather than efficiency is the prime consideration when cyclists choose the "most favourable" gear ratios, and thus pedalling rates for a particular competition. The pedalling rates used in the training system reported here were based on these considerations.

Exercise Heart Rates. The exercise heart rates of cyclists competing in road time trials have been reported to be in the range 170 - 195 beats/min (Brooke and Davies 1971). This highlights the intensity of the load placed upon the cardiorespiratory system, and the need to replicate this in training situations. The heart rate ranges used in this study were based on these research findings.

Conclusions
Indoor simulated cycling training appears to be a valid training method which complies with the principle of the specificity of training. If the system is calibrated it permits measures to be taken of the work load imposed upon the subjects during each training session, for the work load to be modified according to individual needs, and for it to be compared with race performances during the same period.

The use of intermittent work training methods allows the subjects to tolerate greater work loads than normally met in the course of competition. Thus the training method follows the principle of progressive overload

References
Astrand, P.O. and Rodahl, K (1970): Physical training. Textbook of Work Physiology.

Brooke, J.D., Hamley, E and Stone, P (1970): A review of the physical and mental state variables that predict ability at bicycle time trial riding. Proceedings of the World Congress of Sports Medicine. British Journal of Sports Medicine.

Brooke, J.D., and Davies, G.J (1971): The energy requirements of a sports group carrying out prolonged severe exercise over hours. Glucose Ingestion and Physical Work, Salford University, 2-3.

Brooke, J.D. and Firth, M.S. (1974): Calibration of a simple eddy current ergometer. British Journal of Sports Medicine, 8, 2/3, 120-125.

Ekblom, B, Astrand, P.O., Saltin, B, Stenbourg, J., and Wallstrom, B. (1968): Effects of training on circulatory response to exercise. Journal of Applied Physiology, 24, 518-528.

Fox, E.L. (1984): Sprint and endurance training: methods and effects. Sports Physiology, Holt-Saunders, 202-205.

Gleser, M.A., and Vogel, J.A., (1971): Endurance effects of work-rest schedules. Journal of Applied Physiology, 31, 735-739.

Hamley, E.J., and Thomas, V. (1967): Physiological and postural factors in the calibration of the bicycle ergometer. Proceedings of the Physiological Society, April, 55-56.

Nett, T. (1970): Physiological basis of running. Run, Run, Run. 32, 222-223.

Pickering, R (1972): Strength training for athletes. British Amateur Athletic Board, London.

Rash, P.J., and Morehouse, L.E. (1957): Effect of static and dynamic exercise on muscular strength and hypertrophy. Journal of Applied Physiology, 11, 29.

Rowell, L.B., (1964): Limitations of predictions of VO2max. Journal of Applied Physiology, 19, 5, 919-927.

Ulmer, H.V. (1973): The pedalling speeds of racing cyclists participating in track racing and ergometer trials. Sportarzt und Sportmedizin, 4, 77-82.

Watson, A.W.S. (1983): Features of training: warm-up, motor unit types. Physical Fitness and Athletic Performance, London, 72.

Whitt, F.R. (1971): A note on the energy expenditure of sporting cyclists. Ergonomics, 14, 3, 419-424.