Some of the most important problems of the calculus are those where time is the independent variable, and we have to think about the values of some other quantity that varies when the time varies. Some things grow larger as time goes on; some other things grow smaller. The distance that a train has got from its starting place goes on ever increasing as time goes on. Trees grow taller as the years go by. Which is growing at the greater rate; a plant \(12\) inches high which in one month becomes \(14\) inches high, or a tree \(12\) feet high which in a year becomes \(14\) feet high?

In this chapter we are going to make much use of the word *rate*. Nothing to do with poor-rate, or water-rate (except that even here the word suggests a proportion—a ratio—so many pence in the pound). Nothing to do even with birth-rate or death-rate, though these words suggest so many births or deaths per thousand of the population. When a motor-car whizzes by us, we say: What a terrific rate! When a spendthrift is flinging about his money, we remark that that young man is living at a prodigious rate. What do we mean by *rate*? In both these cases we are making a mental comparison of something that is happening, and the length of time that it takes to happen. If the motor-car flies past us going \(10\) yards per second, a simple bit of mental arithmetic will show us that this is equivalent—while it lasts—to a rate of \(600\) yards per minute, or over \(20\) miles per hour.

Now in what sense is it true that a speed of \(10\) yards per second is the same as \(600\) yards per minute? Ten yards is not the same as \(600\) yards, nor is one second the same thing as one minute. What we mean by saying that the *rate* is the same, is this: that the proportion borne between distance passed over and time taken to pass over it, is the same in both cases.

Take another example. A man may have only a few pounds in his possession, and yet be able to spend money at the rate of millions a year—provided he goes on spending money at that rate for a few minutes only. Suppose you hand a shilling over the counter to pay for some goods; and suppose the operation lasts exactly one second. Then, during that brief operation, you are parting with your money at the rate of \(1\) shilling per second, which is the same rate as £\(3\) per minute, or £\(180\) per hour, or £\(4320\) per day, or £\(1,576,800\) per year! If you have £\(10\) in your pocket, you can go on spending money at the rate of a million a year for just \(5\frac{1}{4}\) minutes.

It is said that Sandy had not been in London above five minutes when “bang went saxpence.” If he were to spend money at that rate all day long, say for \(12\) hours, he would be spending \(6\) shillings an hour, or £\(3\). \(12\)*s*. per day, or £\(21\). \(12\)*s*. a week, not counting the Sawbbath.

Now try to put some of these ideas into differential notation.

Let \(y\) in this case stand for money, and let \(t\) stand for time.

If you are spending money, and the amount you spend in a short time \(dt\) be called \(dy\), the *rate* of spending it will be \(\dfrac{dy}{dt}\), or rather, should be written with a minus sign, as \(-\dfrac{dy}{dt}\), because \(dy\) is a *decrement*, not an increment. But money is not a good example for the calculus, because it generally comes and goes by jumps, not by a continuous flow—you may earn £\(200\) a year, but it does not keep running in all day long in a thin stream; it comes in only weekly, or monthly, or quarterly, in lumps: and your expenditure also goes out in sudden payments.

A more apt illustration of the idea of a rate is furnished by the speed of a moving body. From London (Euston station) to Liverpool is \(200\) miles. If a train leaves London at \(7\) o’clock, and reaches Liverpool at \(11\) o’clock, you know that, since it has travelled \(200\) miles in \(4\) hours, its average rate must have been \(50\) miles per hour; because \(\frac{200}{4} = \frac{50}{1}\). Here you are really making a mental comparison between the distance passed over and the time taken to pass over it. You are dividing one by the other. If \(y\) is the whole distance, and \(t\) the whole time, clearly the average rate is \(\dfrac{y}{t}\). Now the speed was not actually constant all the way: at starting, and during the slowing up at the end of the journey, the speed was less. Probably at some part, when running downhill, the speed was over \(60\) miles an hour. If, during any particular element of time \(dt\), the corresponding element of distance passed over was \(dy\), then at that part of the journey the speed was \(\dfrac{dy}{dt}\). The *rate* at which one quantity (in the present instance, *distance*) is changing in relation to the other quantity (in this case, *time*) is properly expressed, then, by stating the differential coefficient of one with respect to the other. A *velocity*, scientifically expressed, is the rate at which a very small distance in any given direction is being passed over; and may therefore be written \[v = \dfrac{dy}{dt}.\]

But if the velocity \(v\) is not uniform, then it must be either increasing or else decreasing. The rate at which a velocity is increasing is called the *acceleration*. If a moving body is, at any particular instant, gaining an additional velocity \(dv\) in an element of time \(dt\), then the acceleration \(a\) at that instant may be written \[a = \dfrac{dv}{dt};\] but \(dv\) is itself \(d\left( \dfrac{dy}{dt} \right)\). Hence we may put \[a = \frac{d\left( \dfrac{dy}{dt} \right)}{dt};\] and this is usually written \(a = \dfrac{d^2y}{dt^2}\);

or the acceleration is the second differential coefficient of the distance, with respect to time. Acceleration is expressed as a change of velocity in unit time, for instance, as being so many feet per second per second; the notation used being \(\text{feet} \div \text{second}^2\).

When a railway train has just begun to move, its velocity \(v\) is small; but it is rapidly gaining speed—it is being hurried up, or accelerated, by the effort of the engine. So its \(\dfrac{d^2y}{dt^2}\) is large. When it has got up its top speed it is no longer being accelerated, so that then \(\dfrac{d^2y}{dt^2}\) has fallen to zero. But when it nears its stopping place its speed begins to slow down; may, indeed, slow down very quickly if the brakes are put on, and during this period of *deceleration* or slackening of pace, the value of \(\dfrac{dv}{dt}\), that is, of \(\dfrac{d^2y}{dt^2}\) will be negative.

To accelerate a mass \(m\) requires the continuous application of force. The force necessary to accelerate a mass is proportional to the mass, and it is also proportional to the acceleration which is being imparted. Hence we may write for the force \(f\), the expression

\[\begin{aligned} f&=ma; \\ f&=\dfrac{dv}{dt}; \\ f&= \dfrac{d^2y}{dt^2}. \end{aligned} \]

The product of a mass by the speed at which it is going is called its *momentum*, and is in symbols \(mv\). If we differentiate momentum with respect to time we shall get \(\dfrac{d(mv)}{dt}\) for the rate of change of momentum. But, since \(m\) is a constant quantity, this may be written \(m \dfrac{dv}{dt}\), which we see above is the same as \(f\). That is to say, force may be expressed either as mass times acceleration, or as rate of change of momentum.

Again, if a force is employed to move something (against an equal and opposite counter-force), it does *work*; and the amount of work done is measured by the product of the force into the distance (in its own direction) through which its point of application moves forward. So if a force \(f\) moves forward through a length \(y\), the work done (which we may call \(w\)) will be \[w = f \times y;\] where we take \(f\) as a constant force. If the force varies at different parts of the range \(y\), then we must find an expression for its value from point to point. If \(f\) be the force along the small element of length \(dy\), the amount of work done will be \(f \times dy\). But as \(dy\) is only an element of length, only an element of work will be done. If we write \(w\) for work, then an element of work will be \(dw\); and we have \[dw = f \times dy; \]

which may be written \[\begin{aligned} dw&=ma\cdot dy; \\ dw&=m\dfrac{d^2 y}{dt^2 }\cdot dy; \\ dw&= m\dfrac{dv}{dt}\cdot dy. \end{aligned}\]

Further, we may transpose the expression and write \[ \dfrac{dw}{dy}=f \]

This gives us yet a third definition of *force*; that if it is being used to produce a displacement in any direction, the force (in that direction) is equal to the rate at which work is being done per unit of length in that direction. In this last sentence the word *rate* is clearly not used in its time-sense, but in its meaning as ratio or proportion.

Sir Isaac Newton, who was (along with Leibnitz) an inventor of the methods of the calculus, regarded all quantities that were varying as *flowing*; and the ratio which we nowadays call the differential coefficient he regarded as the rate of flowing, or the *fluxion* of the quantity in question. He did not use the notation of the \(dy\) and \(dx\), and \(dt\) (this was due to Leibnitz), but had instead a notation of his own. If \(y\) was a quantity that varied, or “flowed,” then his symbol for its rate of variation (or “fluxion”) was \(\dot{y}\). If \(x\) was the variable, then its fluxion was called \(\dot{x}\). The dot over the letter indicated that it had been differentiated. But this notation does not tell us what is the independent variable with respect to which the differentiation has been effected. When we see \(\dfrac{dy}{dt}\) we know that \(y\) is to be differentiated with respect to \(t\). If we see \(\dfrac{dy}{dx}\) we know that \(y\) is to be differentiated with respect to \(x\). But if we see merely \(\dot{y}\), we cannot tell without looking at the context whether this is to mean \(\dfrac{dy}{dx}\) or \(\dfrac{dy}{dt}\) or \(\dfrac{dy}{dz}\), or what is the other variable. So, therefore, this fluxional notation is less informing than the differential notation, and has in consequence largely dropped out of use. But its simplicity gives it an advantage if only we will agree to use it for those cases exclusively where *time* is the independent variable. In that case \(\dot{y}\) will mean \(\dfrac{dy}{dt}\) and \(\dot{u}\) will mean \(\dfrac{du}{dt}\); and \(\ddot{x}\) will mean \(\dfrac{d^2x}{dt^2}\).

Adopting this fluxional notation we may write the mechanical equations considered in the paragraphs above, as follows:

distance | \(x\) |

velocity | \(v = \dot{x}\), |

acceleration | \(a = \dot{v} = \ddot{x}\), |

force | \(f = m\dot{v} = m\ddot{x}\), |

work | \(w = x \times m \ddot{x}\). |

*Examples*.

We see that the conditions of the motion can always be at once ascertained from the time-distance equation and its first and second derived functions. In the last two cases the mean velocity during the first \(10\) seconds and the velocity \(5\) seconds after the start will no more be the same, because the velocity is not increasing uniformly, the acceleration being no longer constant.