ple development of the carriage, the omnibus of the
coach, and the supplementary traffic of the underground and electric
railways is a by no means brilliantly imagined adaptation of the
long-route railway. These are all still new things, experimental to the
highest degree, changing and bound to change much more, in the period of
specialization that is now beginning.

Now, the first most probable development is a change in the omnibus and
the omnibus railway. A point quite as important with these means of
transit as actual speed of movement is frequency: time is wasted
abundantly and most vexatiously at present in waiting and in
accommodating one's arrangements to infrequent times of call and
departure. _The more frequent a local service, the more it comes to be
relied upon._ Another point--and one in which the omnibus has a great
advantage over the railway--is that it should be possible to get on and
off at any point, or at as many points on the route as possible. But
this means a high proportion of stoppages, and this is destructive to
speed. There is, however, one conceivable means of transit that is not
simply frequent but continuous, that may be joined or left at any point
without a stoppage, that could be adapted to many existing streets at
the level or quite easily sunken in tunnels, or elevated above the
street level,[11] and that means of transit is the moving platform,
whose possibilities have been exhibited to all the world in a sort of
mean caricature at the Paris Exhibition. Let us imagine the inner circle
of the district railway adapted to this conception. I will presume that
the Parisian "rolling platform" is familiar to the reader. The district
railway tunnel is, I imagine, about twenty-four feet wide. If we suppose
the space given to six platforms of three feet wide and one (the most
rapid) of six feet, and if we suppose each platform to be going four
miles an hour faster than its slower fellow (a velocity the Paris
experiment has shown to be perfectly comfortable and safe), we should
have the upper platform running round the circle at a pace of
twenty-eight miles an hour. If, further, we adopt an ingenious
suggestion of Professor Perry's, and imagine the descent to the line
made down a very slowly rotating staircase at the centre of a big
rotating wheel-shaped platform, against a portion of whose rim the
slowest platform runs in a curve, one could very easily add a speed of
six or eight miles an hour mo