Interstellar space is filled with diffuse gas and dust. Relatively denser and cooler regions, up to 50 pc in diameter and with a million solar masses, are filled with molecules. In these molecular clouds, shock fronts from nearby star formations or a supernova explosion or some other global gravitational disturbance may begin the process of self‐gravitational contraction, leading to the formation of new stars. The earliest stages of pre–main sequence evolution are not directly observable, because protostars are hidden behind massive amounts of dust. Consequently, no radiation from the forming star is visible. If one could envision a protostar without the obscuring dust, theory suggests that initially a protostar would be very cool but luminous, with convection very efficiently moving gravitational energy that is released in the interior outwards to the exterior. As the object shrank, the surface area would dramatically decrease and the overall luminosity likewise decrease rapidly.
Some of these earliest stages of evolution are believed to occur in the small, dense, dark dust clouds that often are seen silhouetted against more extended regions of luminous, hot, interstellar nebulae. These are the Bok globules. Observation of radio radiation that penetrates the dust from these sites suggests that internal motions of the interstellar material are in a stage of contraction. Such an object may also be termed a cocoon star, because of the surrounding shroud of dense, opaque dust. When the dust is sufficiently warmed by radiation from the interior protostar, it in turn will radiate in the infrared. Many infrared sources are observed in regions where star formation is taking place. This stage of evolution is also termed Helmholtz contraction—one‐half of the energy released by gravitation contraction into the protostellar material results in heating, and one‐half of the energy is convected to the surface to be radiated away.
As the core temperature of the protostar rises, ionization of the material occurs. Photons are not absorbed as well by ionized mate‐rial, thus a transparent radiative core forms in which the energy is transported by photons. Photons, however, cannot directly move to the surface because they are continually colliding with the nuclei and electrons. In a collision, a photon's direction is changed; it is just as likely to be reflected back into the interior as not. Photons slowly drift outwards to the surface, but along the way each undergoes a tremendous number of collisions, and the time to ultimate escape is large, amounting to as long as a million years or more. At this stage the luminosity reaching the surface has declined greatly, but now starts to slowly rise as the protostar continues its contraction. The protostar's surface temperature also increases, thus it now moves parallel to the main sequence in the HR diagram (see Figure ). When the central temperature rises to about 10 6 K, the first energy generation via nuclear reactions commences. By this time the outer layers of surrounding material may be blown away, revealing a new star.
HR diagram showing pre–main sequence evolution for stars of different masses. The dashed line separates those stages (at right) that are hidden within dense clouds of dust and thus not directly observable from those stages (at left) that can be directly seen when the surrounding dust has been blown away and dispersed.
A star does not become stable instantaneously, however. By analogy, imagine a rubber ball sitting on a table. The ball is in a state of equilibrium, with the surface tension of the tabletop pushing up on the ball to balance the downward gravitation force exerted by Earth. When dropped from a height, the ball bounces around for a while before that final equilibrium is reached. A new star is similar: Gravitation contracts the star, heating the central material until thermonuclear reactions begin. But at this point, the star is overcontracted. The new source of energy produces sufficient heating to increase central pressure and overbalance gravity. Thus the central core begins to expand, but expansion is accompanied by a cooling that damps the thermonuclear reactions. Now pressure is too low to balance gravity, and core contraction begins anew. This “bouncing” of the central core around its ultimate state of equilibrium is matched by changes in the gravitation‐pressure balance in the outer part of the star. The surface responds to variations in the core with erratic variations in the surface radius, temperature, and luminosity, producing a T Tauri variable star. FU Orionis is an example of such a young variable star. FU Orionis was discovered in 1939 only when its surrounding dust became sufficiently transparent for its light to be seen.
In general properties, T Tauri stars are young, typically 10 5 to 10 6 years old, and usually found in widely dispersed groups ( T associations) embedded in regions of gas and dust (for example, in the direction of the Orion Nebula). Their brightnesses and surface temperatures place them to the right of the main sequence (see Figure ). They are erratically variable, with amplitudes up to 3 magnitudes (a factor of 16 between minimum and maximum brightness). Their spectra show intense emission lines from extensive outer coronas indicating substantial mass loss in strong stellar winds at velocities of 70 to 200 km/s. Star formation is not a very efficient process in converting interstellar material into new stars.
T Tauri stars in the HR diagram.
Only relatively low mass stars are observed in their T Tauri stage. Higher‐mass stars with their greater gravity pass through this stage so quickly that the likelihood of actually observing a high‐mass T Tauri star is very small. A protostar on its way to becoming a 17 solar mass B0 star, for example, may take as little as 100,000 years to contract and achieve main sequence stability. A star like the Sun contracts much more slowly, taking about 30,000,000 years to become a full fledged main sequence star, whereas a low‐mass M5 star (0.2 solar masses) could take 500,000,000 years to accomplish its main sequence stage of stability.
Related to T Tauri stars are the Herbig‐Haro objects. Originally discovered as compact regions of emission from hot interstellar gas, these were soon found to occur in pairs, on opposite sides of an often unseen star in its stage of becoming a main sequence star. Mass loss often occurs in the form of a bipolar outflow from protostars, two jets in opposite directions. When these jets collide with denser regions in the surrounding interstellar material, the gas is heated and can be observed. Herbig‐Haro objects were initially thought to be very young stars, but observations made using the Hubble Space Telescope in 1999 show bipolar outflows in stars that appear to be forming planets, which usually happens when the star is much older. The star HD163296 is one such object; it shows bipolar jets as well as evidence of planetary formation.