The resistance of metallic conductors increases with temperature. The small current running through it (supplied, as you say, by the meter's battery) is heating the tungsten filament. The resistance increases with temperature due to interactions between the electrons and phonons (the general excited state of the atoms or molecules of the substance). It's easy to picture, that as the temperature of the substance goes up, so does the excitement of its atoms or molecules, and so there are more interactions between the phonons and the things that carry current. So -- increased resistance! The description of phonons gets really deep into quantum physics but the concept isn't too hard I think.
Mark: Many thanks. So, indeed, a lightbulb is a terrible dummy load! But 1968 Novice Ham wasn't expected to spend non-existent dollars on a high-power, non-inductive resistor. So, as explained in "How to Become a Radio Amateur," my little transmitter should be loaded up as follows:
First of all, remember that antenna socket J2 is a female RCA (phono) socket.
"Insert a No. 6 screw about 1 inch long into J2 until it makes firm contact. Connect a 15-watt, 115-volt lamp to the output of the transmitter, with one terminal of the lamp connected to the screw at J2, and the other terminal of the lamp connected to some convenient point on the chassis.... If a socket with short leads isn't available for the lamp, solder wires to the shell and base contacts."There you have it, the ham homebrewing philosophy in a nutshell. Don't spend a few cents on an RCA plug and a lamp socket. Instead, ram a No. 6 screw into the antenna socket and solder the wires directly to the bulb, unless of course you happen to have a socket "available" in your (non-existent) Novice junkbox!
Proving that no homebrewed project is ever "finished," however basic the project, this morning I fine-tuned my incandescent dummy load. I had noticed that the shell of the bulb was connected to the center wire of the coax, and the base of the bulb was connected to the braid. This theoretically meant that the "hot" side was connected to a larger expanse of exposed metal (the bulb shell or screw). So this morning I reversed the contacts, just in case it slightly lowers the risk of RFI. (OTOH both legs of a half-wave dipole radiate, don't they? So does it matter which way round the bulb is wired?)
Martin's First Law of Homebrewing: If you connect a two-wire lead to something, and do it at random, you will always connect it the "wrong" way round the first time even though the statistical probability of getting it right is actually 50 percent!
73 de Martin, KB1WSY