As mentioned above, the critical day length is 15 hours in Pharbitis nil and it is 12 hours in Sinapis alba. In early days when photoperiodism was found, it was thought that the length of day is important. Later, however, it was clarified that the length of night is important. That is, a short-day plant flowers when the night length is longer than a certain length, and a long-day plant flowers when the night length is shorter than a certain length. Such a critical length of night is called critical night length. The critical night length in P. nil is 9 hours, and the critical night length in S. alba is 12 hours.
　The photoperiodic responses of some plant species are quite accurate. Xanthium strumarium continued vegetative growth when night length was 8 hours and 15 minutes while all the plants flowered when night length was 9 hours. Oryza sativa and Perilla frutescens can distinguish the difference of even 15 minutes.
　The number of photoperiodic cycles required for flowering is different among plant species. Chrysanthemum moriforium, a short-day plant, requires a few suitable photoperiodic cycles. X. strumarium and P. nil, both short-day plants, and Lolium temulentum, a long-day plant, are more sensitive to photoperiodic condition, and can flower even when exposed to only a single suitable photoperiodic cycle. Once plants are induced to flower under the suitable photoperiodic condition, the flowering process would proceed even after they are transferred to a non-suitable photoperiodic condition.

　　The night period regulating photoperiodic flowering must be continuous. The effect of the night period is cancelled when the night period is interrupted by light. That is, a light period inserted in the middle of a long night inhibits flowering of short-day plants, and promotes flowering of long-day plants. This effect of the inserted light is called night break.
　　Even a brief irradiation, sometimes only for a few minutes, is effective as a night break.　The most effective light for the night break is red light. The effect of red light is cancelled when followed by far-red light. This means that the measurement of night length involves phytochrome. The far-red light-absorbing form of phytochrome（Pfr）synthesized during the light phase is changed to the red light-absorbing form （Pr）or is degradated during the dark period. When the amout of Pfr decreased to a certain level, flowering would be induced in short-day plants. Red light given in the middle of the night phase, changs Pr to Pfr and the amount of Pfr returns to the level inhibiting flowering of short-day plants. Far-red light given after red light changes Pfr to Pr and nullifies the effect of a night break.

　　The amount of Pfr decreased to an undetectable level after 1 to 2 hours from the start of the dark phase in Pharbitis nil. Since this is much shorter than the critical night length for P. nil which is 9 hours, the disappearance of Pfr itself does not explain the induction of flowering. The decrease in Pfr level has been explained to start a biological clock that measures the length of night. However, details about the biological clock, and how it measures time remain unknown.
　　In the 8-hour light and 16-hour dark regime, the sensitivity to night break is highest at the 8th hour of the dark period, and low before and after this point. The sensitivity to night break changes periodically when the dark period is longer than 24 hours. That is, the most sensitive phase comes every ca. 24 hours. This circadian rhythm indicates the involvement of a biological clock in the photoperiodic flowering.

　　Short-day plants are induced to flower when their leaves are covered with light-proof bags for a period of time longer than the critical night length. The shoot apical meristem where a flower bud is formed does not need to be exposed to an inductive photoperiodic cycle. Therefore, the photoperiodic stimulus inducing flowering is perceived by leaves.
　　Only mature leaves respond to a photoperiodic signal in many plant species, whereas in Pharbitis nil and Chenopodium rubrum cotyledons have high sensitivity to a photoperiodic signal. The sensitive plants such as P. nil, Xanthium strumarium and Perilla frutescens could be induced to flower by exposing a single leaf to a suitable photoperiodic cycle. The minimum area of a leaf or a cotyledon sufficient for inducing flowering is 1 cm2 in P. nil and Perilla frutescens.

　　The site of photoperiodic perception is the leaf and the site of floral bud formation is the shoot apex. This means that some information must be transmitted from the leaf to shoot apex. This information is called flowering stimulus. The transmission of information in plants is generally due to movement of plant hormone-like substances. From these facts, Chailakhyan proposed the florigen (flowering hormone) theory in 1937. Florigen is generated in leaves exposed to a suitable photoperiodic cycle, transmitted through phloem, and activates the flowering genes at the shoot apex.
　　The biochemical nature of florigen is still unknown. Currently, the existence of florigen can be presumed only by grafting experiments. One of the most important goals in the study on flowering has been to isolate and characterize florigen .

　　When a leaf of a flowering plant of Maryland Mammoth tobacco, a short-day plant, was grafted to a plant of Nicotiana sylvestris, a long-day plant, in a vegetative state, the latter was induced to flower under short-day conditions. Alternatively, grafting of photoinduced leaves of N. sylvestris to a tobacco plant in a vegetative state induced flowering of this short-day tobacco under long-day conditions. If leaves of the stock plant (donor) were removed in this grafting, the scion (receptor) remained vegetative. This indicates the existence of a universal florigen that is not species-specific.
　Among the grafting experiments, the induction of flowering in the plant species which does not flower under ordinal condition is convincing. Ipomoea batatas is a short-day plant, but its sensitivity to a short-day condition is very weak, and therefore it is hard to flower. It never flowers in the seedling stage under laboratory conditions. However, I. batatas seedlings grafted to Pharbitis nil seedlings are induced to flower quite easily by short-day treatment. This indicates the existence of florigen.

　　The original article where Chailakhyan proposed his florigen theory is difficult to access because it was written in Russian for the Academy of the Soviet Union in the early 20th century. Many plant physiologists today probably have not read his original article on florigen, and the florigen concept is now understood in several different ways resulting in confusion. It has been misunderstood that florigen was proposed as a single component that is common in the plant kingdom.
　　The grafting experiments showed that the flowering stimulus could be transmitted between different plant species suggesting that florigen was common among species. Although gibberellins can induce flowering of some long-day plants under a noninductive condition, gibberellins are not considered as florigen because they do not induce flowering of short-day plants. However, there is no reason to consider that florigen needs to be ubiquitous. The florigen of long-day plants could be a gibberellin and the florigen of short-day plants might be a chemical other than a gibberellin. The results of grafting experiments mentioned above should not be generalized because grafting is possible only between taxonomically close relatives.
　　Ferns have a regulation mechanism of reproduction similar to that of flowering plants. The formation of antheridium, a male reproductive organ, is induced by an antheridiogen produced by gametophytes. Antheridiogens are distributed commonly in ferns, but they do not have a common structure. They are almost species-specific, and similar within a closely-related taxonomic group. Considering this fact and the similarity between antheridiogen and florigen, it is not necessary to assume that florigen is a single chemical commonly active in all plants.

　　In short-day plants like Pharbitis nil, florigen acts at the shoot apex to induce flower formation. In rosette-forming long-day plants, on the other hand, there are some additional phenomena before the flower formation, that is, stem elongation, cauline leaf formation, and inflorescence formation.
　　Stem elongation and inflorescence formation in Lolium temulentum, a long-day plant, are induced by different kinds of gibberellins. Inhibitors of gibberellin synthesis inhibit stem elongation but not flower formation in spinach. Thus, stem elongation and flower formation are regulated by different factors. The florigen of long-day plants may promote both stem elongation and production of another factor which directly induces flower formation. Another possibility is that a long-day condition promotes stem elongation through gibberellin synthesis on one hand and induces flower formation through florigen synthesis on the other hand. If long-day plants and P. nil have a common florigen, and the florigen of P. nil could be tested on long-day plants, which would be induced stem elongation or flower formation?

　　In the grafting experiments with Perilla crispa by Zeevaart in 1962, a photoinduced leaf grafted onto vegetative stocks continued to function as the donor of florigen through several repeated graftings. Therefore, florigen was considered to be active for a long time. However, the system producing florigen in the photoinduced leaves may continue to function, and it may continue to produce florigen. If this were the case, the longevity of florigen could not be estimated.
　　In the experiment with Ipomoea batatas grafted onto Pharbitis nil, the florigen of P. nil was estimated to be active for 4 to 6 days.

　　A photoinduced leaf excised from flowering Xanthium strumarium, a short-day plant, and grafted onto vegetative X. strumarium, induced the latter to flower under a long-day condition. Also, a leaf excised from the latter plant (never exposed to a short-day condition) and grafted onto another vegetative plant could act as a doner leaf to induce flowering. This is called indirect or secondary induction. This indicates that florigen itself triggers the production of florigen.