In 1934 kulIna kAyastha bengali people aged 20-30 years had generation numbers between 23 and 30 inclusive from an arbitrary starting point (see the tradition page), with the majority between 26 and 29 inclusive. (This hold also for brahmins: Roughly speaking, my generation number should be between 25 and 32, probably between 28 and 31. It is actually 31). Thus, if the last 700 years or so have been generic, one expects a fluctuation of about 10%.
In a particular family, 1. rAmasantoSa was young around 1724 AD 2. ratneshvara 3. gurudAsa 4. kAlidAsa 5. candrashekhara 6. shashishekhara 7. mrgAnkabhUSaNa was young in 1934 and this is a chain along eldest male heir. This would give a generation time of 35 years. However, over a long stretch it could easily be 28 years.
In the mogala rulers 1. bAbara (b. 1483 k. -1530) 2. humAyuna (k. 1530-1556) 3. Akbara (b. 1542 k. 1556-1605) 4. jAhAngIra (k. 1605-1657) 5. shAjAhAna (k. 1628-1658) 6. Arangajeba (k. 1658-1707) 7. bAhAdura-shA (b. 1643 k. 1707-1712), the generation time from bAbara to bahadur-shA is roughly 26 years, whereas the mean kingship time from humAyuna to him was 35.4 years.
In the English rulers 1. William I (1066-1087) 10. Edward II (1307-1327) 19. Henry II (1485-1509) 28. James II (1685-1688) 37. Edward II (1901-1910) shows that the mean kingship time varies (26.1, 20.2, 20.3 and 22.5 in the intervals) giving an overall mean of 22.8 years. This is slightly lower than what is expected of generation times because sometimes the next king was not the son of the previous king. Similarly Richard II to Mary is 11 monarchs between 1377 and 1558. This gives a mean of 16.4: but in this period 6 times in 11 monarchs, the flow of succession was interrupted.
girIndrashekhara actually did a survey in the first half of the 20th century of some students with data from the student welfare commitee of calcutta university. They find that combining the kAyastha and brAhmaNa data, the first male child is usually born at 27.16 +/- 0.19 years of father's age (sample size 403, sd 5.75), second at 30.36 +/- 0.18 (401, 5.47) and third at 33.79 +/- 0.22 (359, 6.41).
In the British Registrar General's Data C.R. Rich: "The measurement of the rate of population growth" Journal of the Institute of Actuaries, Vol. LXV, part no 311, 1934, Table 5, p. 52, we find that the mean age of the mother at the birth of her first daughter is 28.9 (1861 AD-1870), 29.0 (1871-1880), 29.3 (1881-1890), 29.6 (1891-1900), 29.9 (1901-1910), 30.0 (1910-1912), 29.8 (1920-1922).
So, one can conclude that a generation time above 25 years is the norm rather than the exception. In shoirt stretches of 7-10 generation, numbers as high as 35 are also quite common.