Sky & Telescope - September 2020 - 28

Art and Astronomy

A Curious Clock and
Missing Bells
A close look at View of Delft reveals
another remarkable clue. The façade of
the building known as the Schiedam
Gate includes a clock! However, a blob
of paint on the lower left of the clock
face makes it difﬁcult to read the exact
time. We surveyed the existing literature and found that previous Vermeer scholars have interpreted the indicated time in
a variety of ways: "a time between seven and half past seven,"
"just past 7 o'clock," "ten past seven," "exactly 7:12," and
"about 7:15 to 7:30 A.M."
Some of these authors, especially those offering precise
times, appear to make the incorrect assumption that the
Schiedam Gate clock possessed both an hour and a minute
hand. However, minute hands were not employed on tower
clocks until the late 19th century. Tower clocks in the 17th
century always had only a single hour hand, according to independent studies by G. Roosegaarde Bisschop, Laura MeilinkHoedemaker, and Wim Weve, three experts on Dutch architecture. Many prints of Delft from the 17th and 18th centuries
include the Schiedam Gate and its clock, always with a single,
q NAME THE HOUR This detail, from just left of center in View of Delft,
shows the Schiedam Gate clock. A globule of paint at the lower left of
the clock face makes the displayed time difficult to read. Tower clocks
in the 17th century had only a single long hour hand, with the front half
pointing to the time and the back half acting as a counterweight. The
hand's straight line, extending here from upper right to lower left, appears to indicate a time near 8:00 a.m.

SEPTEMBER 2 02 0 * SK Y & TELESCOPE

t CARILLON Vermeer's View of Delft shows
an empty bell tower at the octagon level of
the Nieuwe Kerk and therefore dates from
before the installation of the carillon and clock
faces seen in this detail from a 1667 print
by Coenraet Decker, with the view looking
northeast from Delft's marketplace. François
Hemony began installing the carillon in the
tower during the spring of 1660. The column
of four large bells on the right side of this view
would have been visible from Vermeer's location south of the church.

long hour hand. The front half of the
hand points to the time, while the back
half acts as a counterweight.
In our analysis of Vermeer's painting we noted that the
clock's hand extends on the clock face from upper right to
lower left, suggesting that the time depicted is near 8:00
a.m. The alternate interpretation, of 2:00 p.m., is ruled out
because the sunlight is coming from the south-east, indicating a morning clock time.
Today, four sides of the Nieuwe Kerk octagon feature clock
faces, and three sides host the bells of a carillon. These bells
offer another important clue. According to essays about the
Delft carillon by Laura Meilink-Hoedemaker on her website,
and by Adelheid Rech on the Essential Vermeer website, the
Dutch foundry of François Hemony cast the carillon bells in
1659 and 1660. Hemony installed 20 bells beginning in April
and May 1660, followed by 10 bells in July 1660, and the last
three bells in August and September 1660.
As art historians have noted, Vermeer's painting shows
no bells in the octagon. The canvas therefore must depict the
situation before the installation of the carillon in 1660.

An April or September Sun?
We calculated the Sun's position in the sky of Delft at 8:00
a.m. on each day for 1659, with dates in the Gregorian calendar system, which South Holland adopted in 1583. Because
modern time zones did not exist in the 17th century, we calculated the position of the solar disk on each day at 7:43 UT,
equivalent to 8:00 a.m. local mean time in Delft. The requirement that the Sun's azimuth must be near 110° to produce
the shadows depicted on the Nieuwe Kerk tower was met on
only two dates: April 6th and September 3rd. We also did the
same calculations for 8:00 a.m. apparent solar time (the time
on sundials, which were possibly used to set the mechanical
clocks), with the corresponding dates shifting slightly to April
8th and September 4th. These April dates can be ruled out
as being too early in the year, because the painting depicts
abundant leaves on the trees, something that does not occur
until late April or May at this latitude. That means September
3rd or 4th are the likeliest dates.
This September result conﬂicts with the previous consensus that View of Delft depicts the scene in the late spring or
early summer, not far from the summer solstice. However,
the Sun then rises in the northeast, which means the octagon

PAIN TING DE TAIL: JOH A NNES V ER MEER / M AURITSHUIS, THE
H AGUE; CLOCK ENG R AVING: RIJKSM USEU M, A MSTERDA M

eral weeks during the autumn of 2019.
The shadows on the Nieuwe Kerk in
our photos matched those in Vermeer's
painting only when the Sun's azimuth
was near 110°.

Sky & Telescope - September 2020

Table of Contents for the Digital Edition of Sky & Telescope - September 2020