The size of the flame should fit the size of the candle, of course. At least that is the first guess of most; common knowledge, so to say. That is not the case here, the flames are too big for both sizes of candles, about one third too big but still look kind of right, right?
Oh, and neither the form of the flame nor the form of the wick are correct, the pencil sketch at the upper right comes a bit closer to the correct shape. Most significant differences are the roundish tip of the flame instead of the pointy one common in illustrations and the curved wick.
The curve in the wick is very famous invention that made the wick shorten itself and saved the people of the times before the widespread electrification a lot of the tedious wick-trimming.
Jules Léonard Louis Cambacérès did a detailed description of the curved wick. He did not invent it as he mentions a patent in “Brevets expirés” volume 41 of which I was not able to find a copy online (neither did I find a copy of Cambacérès article, only a German Translation of it).
I made the sketches in watercolor and it seems to be good style to list the colors used (probably because the chain of image processing is rarely fully color-corrected. This picture for example has been taken with a cheap digi-snapper). The colors used are: pure cadmium yellow and -red for the flames with a bit of cobalt blue at the bottom (cobalt blue is too light, you may give ultramarine a try for your own version), pains grey (blueish) with a drip of cobalt blue for the candles and pure for the wicks.
The tip of the wick glows; I decided to ignore that but if yu want to add it I would propose a bit of orange with a little point of darkened rose madder or alizarin in the middle. The Wikipedia page about candles has a nice picture showing the glowing tip of the wick..