Table 1

Fitting terms from least-squares linear regressions of log10 maximum force output (N) as a function of log10 motor mass (kg) for different types of motors

Motor typeInterceptSloper2NSE slopeMinimum massMaximum mass
Group 1
 Muscles2.9950.6770.997160.0104.8 × 10−220.014
 Rockets2.7370.7190.988190.0190.00795.87 × 105
 Linear actuators3.0480.6120.65690.1670.342,587
Mean 3.008 0.686
Group 2
 Running animals1.7030.9490.99850.0230.0000432.7
 Swimming animals1.6720.9240.99280.0330.00090.04
 Flying birds1.5550.9590.993110.0260.0020.37
 Flying bats1.8621.0820.99470.0370.0020.013
 Flying insects1.5830.9590.9821490.0113.7 × 10−70.008
 Electric rotary motors1.7391.0780.999100.0130.04558
 Linear induction motors1.8250.8540.94070.0961.932.6
 Piston engines1.7191.0160.988310.0210.1652,744
Mean 1.726 0.977
  • The slope of these log-log regressions is b in the equation y = axb; the inverse log of the intercept is a. N shows the sample size within each category of motor; SE is the standard error of the least-squares regression slope. Minimum and maximum motor masses (kg) are shown for each category. The data are separated into two groups that are distinguished by slope and intercept. Single molecules that create translational motion (myosin, kinesin, and dynein) are included here with muscle cells and whole muscles (treating muscles alone yields a nearly identical scaling equation). Piston engines greater than 4,400 kg mass are excluded, because they scale differently than smaller piston engines (see text and Fig. 3).