Moreover, it is now clear that not only AGNs but also almost all galaxies have massive black holes in their centres. But they emit a very small amount of energy because the black hole has no gas to swallow. One mystery is to know how such massive black holes have been formed and have grown in the cosmic history. It has recently been revealed that a certain sub-class of AGNs, narrow-line Seyfert 1 galaxies (NLS1s), seem to harbor black holes whose mass is rapidly increasing. By estimating the mean duration (life time) and the gas accretion rate of such AGNs, astronomers from Paris Observatory deduced that the black hole will become larger by one to three orders of magnitude in the narrow-line Seyfert 1 phase, on average. It is likely that all galaxies, and all supermassive black holes, have spent some time in this particular phase, in their lifetime. It is now believed that gas accreting toward the central black holes forms a disk-shape, called an "accretion disk", and that this disk is the ultimate power station of these activities. Many studies have been carried out to reveal the physical conditions of the accretion disk in the immediate vicinity of the black hole. Jets can also be fired out around the inner edge of the disk, just adjacent to the black hole region, at nearly the speed of light.
The accretion rate can be estimated when the mass is known, by measuring the flux supposed to be emitted by the accretion disk, for instance in the optical band. It is then found that the speed of BH growth, which can be roughly estimated by the BH mass divided by mass accretion rate, is very high. In other words, NLS1s are apparently young AGNs, with their BHs growing rapidly. This could explain why NLS1s do not seem to follow the same relationship as the other galaxies between the mass of the black hole and the bulge, their black holes being apparently under-massive with respect to the bulge, which should have grown before. Then, can NLS1-phase be the principal phase of the cosmic BH growth history ? This is not as naive a question as it may sound, because the degree of the BH growth completely depends on the duration, or the life time, of this high-accretion rate phase, which has not been clear at all.
As a result of recent, intense observational programs, including ROSAT All Sky Survey and Sloan Digital Sky Survey, etc., it has been recognized that NLS1s are found in about 10 — 30 per cent of AGNs. The deduced relative fraction varies with the wavelength band used to count the AGNs : in the optical it is about 10 per cent, and it is about 30 per cent in soft X-ray band. These relative fractions, combined with a mean duration of an AGN (which is about a hundred million years), have provided an estimation that an active galaxy passes a NLS1-phase for 10 to 30 million years, while it spends as another sub-class the other 70 to 90 million years. Considering a mean gas accretion rate in NLS1s, it is now deduced that a BH grows by one to three orders of magnitude in the NLS1-phase on average. Contrary, it is also realized that a BH does not grow so rapidly in the rest 70 to 90 millions years, in which the gas accretion rate is not so high. One important point is that the relative fraction and the mean accretion rate of NLS1s are currently estimated for relatively nearby objects. In the distant Universe namely in the young Universe), the BH growth may have happened at a faster pace than we would expect it nowadays. This will be revealed in the future by further observations and theoretical progresses.