On this page
Why does a crop sown in May and the same crop sown in June flower in the same week? Why do some plants bloom only in winter in a greenhouse? Two scientists at the US Department of Agriculture answered these questions with a wooden shed, some rail tracks and a lot of patience. This post walks through their reasoning, the numbers from their own paper, and what students can borrow from their method.
What puzzle were Garner and Allard trying to solve?
Wightman W. Garner and Henry A. Allard worked in the Bureau of Plant Industry. In their 1920 paper in the Journal of Agricultural Research they describe two odd plants.
The first was a giant tobacco noticed in 1906 in a strain of Maryland Narrowleaf. Known as Maryland Mammoth, it kept growing until cold weather near Washington, D.C., without ever flowering. Yet seedlings grown in a winter greenhouse always flowered while still small. The second puzzle came from soybeans: certain varieties planted at different dates through the summer tended to flower at about the same date.
How did they rule out the other suspects?
Their paper reads like a detective story. Each suspect was tested and dropped:
Temperature. Mammoth tobacco and soybeans still showed the "winter effect" when the greenhouse was kept as warm as summer outdoors.
Light intensity. In earlier experiments on soybean oil, changing light intensity did not materially change the date of flowering, and partial shading had no clear effect on Mammoth tobacco.
Nutrition. Starving the tobacco in small pots during summer did not make it flower.
That left the length of the daily exposure to light. They admitted it hardly seemed likely, but they tested it anyway.
How did the dark house work?
The first trials began in July 1918 with a small ventilated dark chamber in the field. In 1919 they built a dark house 30 feet by 18 feet, painted black inside, with ventilation that let air in but kept light out. Plants rode on trucks along four steel tracks, so they could be wheeled in each afternoon and out each morning. Control plants stayed outside. Thermographs showed the inside was not colder than outside, so any autumn-like response could not be blamed on cool nights.
What did the soybeans show?
Here are the paper's own figures for the late-maturing Biloxi soybean given light only from 10 a.m. to 3 p.m., a five-hour day. Plants were sown on 8 May, came up on 17 May and went into the dark house on 20 May.
Test plants flowered on 16 June. From emergence that is 14 days left in May plus 16 days in June, so 30 days.
Controls flowered on 4 September. From 17 May: 14 + 30 (June) + 31 (July) + 31 (August) + 4 = 110 days.
Shortening the day brought flowering forward by 110 − 30 = 80 days.
Test plants were 6 to 7 inches tall at flowering, about 15 to 18 cm. Controls were 57 to 58 inches, about 145 to 147 cm. Comparing midpoints, 57.5 ÷ 6.5 ≈ 8.8, so the controls were almost nine times taller.
A gentler treatment told the same story. Biloxi plants given light from 6 a.m. to 6 p.m. came up on 16 June and flowered on 14 July, 28 days later, while controls took until 8 September, 84 days. Soybean observations covered 20 to 25 plants per test.
What did they conclude, and where did the word come from?
The paper's summary states that the relative length of day is a factor of first importance in growth and particularly in sexual reproduction. Some species flower in response to long days, others to short days, and still others under all day lengths found at Washington. Light intensity, from full sunlight down to a third or a quarter of it, was not a factor of importance for flowering.
Two years later, in a short note in Science (1922), they named the response photoperiodism, crediting O. F. Cook of the Bureau of Plant Industry with suggesting the term.
What came after Garner and Allard?
A 2021 review in Frontiers in Plant Science traces the next century: the grouping of plants into long-day, short-day and day-neutral types; Knott's 1934 finding that leaves, not the shoot tip, sense day length in spinach; the observation that a short burst of light during the night can stop short-day plants from flowering; and the search for florigen.
That leaf result answers an exercise that still sits at the end of NCERT Class 11, Chapter 13: would a defoliated plant respond to a photoperiodic cycle? Note that the rationalised chapter (Reprint 2026-27) no longer has a photoperiodism section, and the NEET-UG 2026 syllabus lists plant growth topics without it. Treat this as background unless your board or exam syllabus names it, and check the current year's version.
Common mistakes:
Assuming warmth or brightness triggers flowering. Garner and Allard tested both and found day length mattered.
Forgetting the controls. Every claim above compares test plants with untreated plants sown the same day.
Counting days carelessly. Count from emergence, and remember May has 31 days.
What can a student take from Garner and Allard?
They kept a written list of suspects and crossed them off with evidence, then built equipment just to vary the one factor left. You can use the same habit on a hard problem: list every possible cause, test the cheapest first, and keep a control. If you want help planning which gaps to test first, Fermi's AI tutor can work out what you already know and build a plan for the session.
Next step: recalculate the 6 a.m. to 6 p.m. Biloxi figures yourself and write one sentence on what the 56-day gap tells you.
